Release v3.0.2

feat(tui): queue editing + update_input method
## Queue Editing - Added update_input method to QueueScheduler for editing queued runs - Added update_input method to QueueManager - Added handler for QueuePanelWidget.EditRequested in TUIApp - Only queued (not running) items can be edited ## Queue-First Architecture Support - Queue items can be edited before processing - EditRequested message already existed in QueuePanelWidget 31 tests passing
2026-11-09 03:39:37 -10:00 · 2026-01-02 04:33:19 +00:00 · 2026-01-02 04:24:10 +00:00 · 2026-01-02 04:22:26 +00:00 · 2026-01-02 04:02:55 +00:00 · 2026-01-02 04:01:25 +00:00
56 changed files with 9512 additions and 115 deletions
@@ -16,7 +16,7 @@ RUN mkdir -p /root/.praison
 # Install Python packages (using latest versions)
 RUN pip install --no-cache-dir \
    praisonai_tools \
-    "praisonai>=2.9.2" \
+    "praisonai>=3.0.2" \
    "praisonai[chat]" \
    "embedchain[github,youtube]"

@@ -20,7 +20,7 @@ RUN mkdir -p /root/.praison
 # Install Python packages (using latest versions)
 RUN pip install --no-cache-dir \
    praisonai_tools \
-    "praisonai>=2.9.2" \
+    "praisonai>=3.0.2" \
    "praisonai[ui]" \
    "praisonai[chat]" \
    "praisonai[realtime]" \
@@ -16,7 +16,7 @@ RUN mkdir -p /root/.praison
 # Install Python packages (using latest versions)
 RUN pip install --no-cache-dir \
    praisonai_tools \
-    "praisonai>=2.9.2" \
+    "praisonai>=3.0.2" \
    "praisonai[ui]" \
    "praisonai[crewai]"

@@ -0,0 +1,203 @@
+# Profiling CLI
+
+PraisonAI provides unified profiling capabilities for performance analysis and debugging, with **full visibility from ENTER to first response**.
+
+## Quick Start
+
+```bash
+# Basic profiling
+praisonai "What is 2+2?" --profile
+
+# Deep profiling with call graph, decision trace, and module breakdown
+praisonai "What is 2+2?" --profile --profile-deep
+
+# JSON output format
+praisonai "What is 2+2?" --profile --profile-format json
+```
+
+## Time to First Response (TTFR)
+
+**TTFR** is the key user-perceived latency metric. It measures the time from when you press ENTER until the first visible output appears.
+
+### What contributes to TTFR?
+
+1. **CLI Entry** - Time to reach the CLI entry point
+2. **CLI Parse** - Argparse/Typer routing time
+3. **Routing** - Command routing to handler
+4. **Imports** - Loading required modules
+5. **Agent Init** - Initializing the agent
+6. **Network Request** - Sending request to LLM API
+7. **First Token** - Receiving first token (streaming)
+8. **First Output** - Rendering first visible output
+
+### How to reduce TTFR
+
+1. **Warm runs** - Imports are cached after first run
+2. **Streaming mode** - See output as tokens arrive
+3. **Lighter models** - Faster response times
+4. **Local models** - Eliminate network latency
+
+## CLI Flags
+
+| Flag | Description | Default |
+|------|-------------|---------|
+| `--profile` | Enable profiling with timing breakdown | Disabled |
+| `--profile-deep` | Enable deep profiling with call graph (higher overhead) | Disabled |
+| `--profile-format` | Output format: `text` or `json` | `text` |
+
+## Profile Command
+
+For more detailed profiling options, use the `profile` subcommand:
+
+```bash
+# Profile a query
+praisonai profile query "What is 2+2?"
+
+# With deep call tracing
+praisonai profile query "Hello" --deep
+
+# Limit functions shown
+praisonai profile query "Test" --limit 20
+
+# Save artifacts to directory
+praisonai profile query "Test" --save ./profile_results
+
+# JSON output
+praisonai profile query "Test" --format json
+```
+
+### Profile Query Options
+
+| Option | Description | Default |
+|--------|-------------|---------|
+| `--model, -m` | Model to use | Default model |
+| `--stream/--no-stream` | Use streaming mode | No stream |
+| `--deep` | Enable deep call tracing | Disabled |
+| `--limit, -n` | Top N functions to show | 30 |
+| `--sort, -s` | Sort by: `cumulative` or `tottime` | `cumulative` |
+| `--show-callers` | Show caller functions | Disabled |
+| `--show-callees` | Show callee functions | Disabled |
+| `--save` | Save artifacts to path | None |
+| `--format, -f` | Output format: `text` or `json` | `text` |
+
+## Other Profile Commands
+
+```bash
+# Profile module import times
+praisonai profile imports
+
+# Profile CLI startup time
+praisonai profile startup
+
+# Run comprehensive profiling suite
+praisonai profile suite
+
+# Create performance snapshots for comparison
+praisonai profile snapshot --baseline
+praisonai profile snapshot current --compare
+
+# Configure performance optimizations
+praisonai profile optimize --show
+```
+
+## Output Format
+
+### Text Report
+
+```
+======================================================================
+PraisonAI Profile Report
+======================================================================
+
+Run ID:     abc12345
+Timestamp:  2024-01-02T12:00:00.000000Z
+Method:     cli_direct
+Version:    3.0.1
+
+## Execution Timeline
+---------------------------------------------
+  Imports                   :   500.00 ms
+  Agent Init                :     0.50 ms
+  Execution                 :  2000.00 ms
+  ───────────────────────────────────────────
+  ⏱ Time to First Response  :  2500.50 ms
+  TOTAL                     :  2500.50 ms
+
+## Timing Breakdown
+----------------------------------------
+  Imports:            500.00 ms
+  Agent Init:           0.50 ms
+  Execution:         2000.00 ms
+  ─────────────────────────────────────
+  TOTAL:             2500.50 ms
+
+## Decision Trace (deep profile only)
+----------------------------------------
+  Agent Config:    default
+  Model:           gpt-4
+  Streaming:       disabled
+  Profile Layer:   2
+
+## Top Functions by Cumulative Time
+----------------------------------------------------------------------
+Function                                    Calls   Cumul (ms)
+----------------------------------------------------------------------
+request                                         1      1800.00
+send                                            1      1750.00
+...
+```
+
+### JSON Report
+
+```json
+{
+  "schema_version": "1.0",
+  "run_id": "abc12345",
+  "timestamp": "2024-01-02T12:00:00.000000Z",
+  "invocation": {
+    "method": "cli_direct",
+    "flags": {"layer": 1, "stream": false, "model": null},
+    "praisonai_version": "3.0.1",
+    "python_version": "3.12.0"
+  },
+  "timing": {
+    "total_ms": 2500.5,
+    "imports_ms": 500.0,
+    "agent_init_ms": 0.5,
+    "execution_ms": 2000.0
+  },
+  "functions": [...],
+  "response_preview": "four"
+}
+```
+
+## Profiling Layers
+
+| Layer | Overhead | Features |
+|-------|----------|----------|
+| 0 | < 1ms | Wall-clock phases only |
+| 1 | ~5% | + cProfile function stats |
+| 2 | ~15% | + Call graph (callers/callees) |
+
+- `--profile` uses Layer 1
+- `--profile --profile-deep` uses Layer 2
+
+## Performance Tips
+
+1. **Use Layer 0 for production monitoring** - minimal overhead
+2. **Use Layer 1 for development profiling** - good balance
+3. **Use Layer 2 only for deep debugging** - higher overhead
+
+## Observational Equivalence
+
+The following commands produce equivalent profile reports:
+
+```bash
+# CLI direct
+praisonai "prompt" --profile
+
+# Profile command
+praisonai profile query "prompt"
+```
+
+Both use the same unified profiling architecture and produce identical schema output.
@@ -0,0 +1,257 @@
+# Profiling Module
+
+PraisonAI provides a unified profiling module for programmatic performance analysis.
+
+## Quick Start
+
+```python
+from praisonai.cli.execution import (
+    ExecutionRequest,
+    Profiler,
+    ProfilerConfig,
+)
+
+# Create execution request
+request = ExecutionRequest(prompt="What is 2+2?")
+
+# Configure profiler
+config = ProfilerConfig(layer=1)
+
+# Run with profiling
+profiler = Profiler(config)
+result, report = profiler.profile_sync(request)
+
+# Access results
+print(f"Output: {result.output}")
+print(f"Total time: {report.timing.total_ms}ms")
+```
+
+## ExecutionRequest
+
+Immutable request descriptor for execution.
+
+```python
+from praisonai.cli.execution import ExecutionRequest
+
+# Basic request
+request = ExecutionRequest(prompt="Hello")
+
+# With model
+request = ExecutionRequest(
+    prompt="Hello",
+    model="gpt-4",
+    agent_name="MyAgent",
+)
+
+# With streaming
+request = ExecutionRequest(
+    prompt="Hello",
+    stream=True,
+)
+```
+
+### Parameters
+
+| Parameter | Type | Default | Description |
+|-----------|------|---------|-------------|
+| `prompt` | str | Required | The prompt to execute |
+| `agent_name` | str | "Assistant" | Name of the agent |
+| `agent_instructions` | str | None | Custom instructions |
+| `model` | str | None | Model to use |
+| `stream` | bool | False | Enable streaming |
+| `tools` | tuple | () | Tools to provide |
+
+## ProfilerConfig
+
+Configuration for profiling behavior.
+
+```python
+from praisonai.cli.execution import ProfilerConfig
+
+# Layer 0: Wall-clock only (< 1ms overhead)
+config = ProfilerConfig(layer=0)
+
+# Layer 1: + cProfile stats (~5% overhead)
+config = ProfilerConfig(layer=1)
+
+# Layer 2: + Call graph (~15% overhead)
+config = ProfilerConfig(
+    layer=2,
+    show_callers=True,
+    show_callees=True,
+)
+
+# From CLI flags
+config = ProfilerConfig.from_flags(
+    profile=True,
+    deep=True,
+    network=False,
+)
+```
+
+### Parameters
+
+| Parameter | Type | Default | Description |
+|-----------|------|---------|-------------|
+| `layer` | int | 1 | Profiling layer (0, 1, or 2) |
+| `limit` | int | 30 | Max functions to show |
+| `sort_by` | str | "cumulative" | Sort by: cumulative, time, calls |
+| `show_callers` | bool | False | Show caller functions |
+| `show_callees` | bool | False | Show callee functions |
+| `track_network` | bool | False | Track network requests |
+| `output_format` | str | "text" | Output format: text, json |
+| `save_path` | str | None | Path to save artifacts |
+
+## Profiler
+
+The main profiler class that wraps execution.
+
+```python
+from praisonai.cli.execution import Profiler, ProfilerConfig, ExecutionRequest
+
+config = ProfilerConfig(layer=1)
+profiler = Profiler(config)
+
+request = ExecutionRequest(prompt="Hello")
+result, report = profiler.profile_sync(request, invocation_method="api")
+```
+
+### Methods
+
+#### `profile_sync(request, invocation_method="profile_command")`
+
+Profile synchronous execution.
+
+**Parameters:**
+- `request`: ExecutionRequest - The execution request
+- `invocation_method`: str - How this was invoked (for report)
+
+**Returns:** Tuple of (ExecutionResult, ProfileReport)
+
+## ProfileReport
+
+The canonical profile report with timing and statistics.
+
+```python
+# Access timing breakdown
+print(f"Total: {report.timing.total_ms}ms")
+print(f"Imports: {report.timing.imports_ms}ms")
+print(f"Agent Init: {report.timing.agent_init_ms}ms")
+print(f"Execution: {report.timing.execution_ms}ms")
+
+# Access function stats (layer 1+)
+if report.functions:
+    for func in report.functions[:5]:
+        print(f"{func.name}: {func.cumulative_time_ms}ms")
+
+# Access call graph (layer 2)
+if report.call_graph:
+    print(f"Edges: {len(report.call_graph.edges)}")
+
+# Export formats
+json_str = report.to_json()
+text_str = report.to_text()
+dict_data = report.to_dict()
+```
+
+### Schema Version
+
+The report uses schema version `1.0`. Schema is frozen after release - new fields are always optional with defaults.
+
+### Mandatory Fields
+
+- `schema_version`: str
+- `run_id`: str
+- `timestamp`: str (ISO 8601)
+- `invocation`: InvocationInfo
+- `timing`: TimingBreakdown
+- `response_preview`: str
+
+### Optional Fields
+
+- `functions`: List[FunctionStat] (layer 1+)
+- `call_graph`: CallGraph (layer 2)
+- `network`: List[RequestTiming] (if track_network)
+
+## ExecutionResult
+
+Result of execution with output and metadata.
+
+```python
+# Access output
+print(result.output)
+print(result.success)
+print(result.run_id)
+
+# Check for errors
+if result.error:
+    print(f"Error: {result.error}")
+
+# Duration
+print(f"Duration: {result.duration_ms}ms")
+```
+
+## Data Bounds
+
+Profile data is strictly bounded to prevent memory issues:
+
+| Data Type | Maximum |
+|-----------|---------|
+| Function stats | 1000 |
+| Call graph edges | 5000 |
+| Network requests | 100 |
+
+## Example: Full Profiling Workflow
+
+```python
+from praisonai.cli.execution import (
+    ExecutionRequest,
+    Profiler,
+    ProfilerConfig,
+)
+
+# Configure deep profiling
+config = ProfilerConfig(
+    layer=2,
+    limit=50,
+    show_callers=True,
+    show_callees=True,
+)
+
+# Create request
+request = ExecutionRequest(
+    prompt="Explain quantum computing in simple terms",
+    model="gpt-4",
+)
+
+# Profile execution
+profiler = Profiler(config)
+result, report = profiler.profile_sync(request)
+
+# Print text report
+print(report.to_text())
+
+# Save JSON report
+with open("profile.json", "w") as f:
+    f.write(report.to_json())
+
+# Analyze timing
+print(f"\nTiming Analysis:")
+print(f"  Import overhead: {report.timing.imports_ms / report.timing.total_ms * 100:.1f}%")
+print(f"  Execution time: {report.timing.execution_ms / report.timing.total_ms * 100:.1f}%")
+
+# Top 5 hotspots
+if report.functions:
+    print(f"\nTop 5 Hotspots:")
+    for func in report.functions[:5]:
+        print(f"  {func.name}: {func.cumulative_time_ms:.2f}ms ({func.calls} calls)")
+```
+
+## Invariants
+
+The profiling module guarantees:
+
+1. **Output Equivalence**: Profiled execution produces identical output to non-profiled execution
+2. **Schema Stability**: Schema version 1.0 is frozen; new fields are always optional
+3. **Bounded Data**: All profile data respects maximum bounds
+4. **Zero Overhead When Disabled**: No profiling code runs unless explicitly enabled
@@ -0,0 +1,147 @@
+"""
+Interactive Default Tools Example
+
+This example demonstrates how PraisonAI's interactive modes (TUI and prompt)
+include ACP and LSP tools by default.
+
+Tool Groups:
+- ACP: acp_create_file, acp_edit_file, acp_delete_file, acp_execute_command
+- LSP: lsp_list_symbols, lsp_find_definition, lsp_find_references, lsp_get_diagnostics
+- Basic: read_file, write_file, list_files, execute_command, internet_search
+
+Usage:
+    python interactive_tools_example.py
+"""
+
+import os
+import sys
+
+# Add paths for development
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '../../src/praisonai'))
+sys.path.insert(0, os.path.join(os.path.dirname(__file__), '../../src/praisonai-agents'))
+
+
+def example_get_default_tools():
+    """Example: Get all default interactive tools."""
+    from praisonai.cli.features.interactive_tools import (
+        get_interactive_tools,
+        ToolConfig,
+        TOOL_GROUPS,
+    )
+    
+    print("=== Default Interactive Tools ===\n")
+    
+    # Show tool groups
+    print("Tool Groups:")
+    for group, tools in TOOL_GROUPS.items():
+        if group != "interactive":
+            print(f"  {group}: {tools}")
+    print()
+    
+    # Get all tools
+    config = ToolConfig(workspace=os.getcwd())
+    tools = get_interactive_tools(config=config)
+    
+    print(f"Total tools loaded: {len(tools)}")
+    print(f"Tool names: {[t.__name__ for t in tools]}")
+    print()
+    
+    return tools
+
+
+def example_disable_acp():
+    """Example: Disable ACP tools."""
+    from praisonai.cli.features.interactive_tools import (
+        get_interactive_tools,
+        ToolConfig,
+    )
+    
+    print("=== Disable ACP Tools (--no-acp) ===\n")
+    
+    config = ToolConfig(workspace=os.getcwd(), enable_acp=False)
+    tools = get_interactive_tools(config=config, disable=["acp"])
+    
+    acp_tools = [t for t in tools if t.__name__.startswith("acp_")]
+    print(f"ACP tools after --no-acp: {len(acp_tools)}")
+    print(f"Remaining tools: {[t.__name__ for t in tools]}")
+    print()
+    
+    return tools
+
+
+def example_disable_lsp():
+    """Example: Disable LSP tools."""
+    from praisonai.cli.features.interactive_tools import (
+        get_interactive_tools,
+        ToolConfig,
+    )
+    
+    print("=== Disable LSP Tools (--no-lsp) ===\n")
+    
+    config = ToolConfig(workspace=os.getcwd(), enable_lsp=False)
+    tools = get_interactive_tools(config=config, disable=["lsp"])
+    
+    lsp_tools = [t for t in tools if t.__name__.startswith("lsp_")]
+    print(f"LSP tools after --no-lsp: {len(lsp_tools)}")
+    print(f"Remaining tools: {[t.__name__ for t in tools]}")
+    print()
+    
+    return tools
+
+
+def example_agent_with_default_tools():
+    """Example: Create an Agent with default interactive tools."""
+    from praisonai.cli.features.interactive_tools import get_interactive_tools, ToolConfig
+    from praisonaiagents import Agent
+    
+    print("=== Agent with Default Interactive Tools ===\n")
+    
+    # Get default tools
+    config = ToolConfig(workspace=os.getcwd())
+    tools = get_interactive_tools(config=config)
+    
+    # Create agent with tools
+    agent = Agent(
+        name="InteractiveAgent",
+        instructions="You are a helpful coding assistant with file and code analysis capabilities.",
+        tools=tools,
+        verbose=False
+    )
+    
+    print(f"Agent created with {len(tools)} tools")
+    print(f"Tools: {[t.__name__ for t in tools]}")
+    print()
+    
+    # Simple test - list files
+    print("Testing agent with list_files tool...")
+    result = agent.start("List the Python files in the current directory using list_files tool.")
+    print(f"Result: {result[:200]}..." if len(str(result)) > 200 else f"Result: {result}")
+    print()
+    
+    return agent
+
+
+if __name__ == "__main__":
+    print("=" * 60)
+    print("PraisonAI Interactive Default Tools Example")
+    print("=" * 60)
+    print()
+    
+    # Example 1: Get default tools
+    example_get_default_tools()
+    
+    # Example 2: Disable ACP
+    example_disable_acp()
+    
+    # Example 3: Disable LSP
+    example_disable_lsp()
+    
+    # Example 4: Agent with default tools (requires API key)
+    if os.environ.get("OPENAI_API_KEY"):
+        example_agent_with_default_tools()
+    else:
+        print("=== Agent Example Skipped (no OPENAI_API_KEY) ===")
+    
+    print("=" * 60)
+    print("Examples Complete")
+    print("=" * 60)
@@ -0,0 +1,83 @@
+# PraisonAI Profiling Examples
+
+This directory contains examples for profiling PraisonAI agent performance.
+
+## Prerequisites
+
+```bash
+pip install praisonai
+export OPENAI_API_KEY=your_key_here
+```
+
+## Examples
+
+### 1. Basic Profiling (`basic_profiling.py`)
+
+Demonstrates programmatic profiling of a single query:
+
+```bash
+python basic_profiling.py
+```
+
+### 2. Suite Profiling (`suite_profiling.py`)
+
+Runs a comprehensive profiling suite with multiple scenarios:
+
+```bash
+python suite_profiling.py
+```
+
+### 3. Optimization Demo (`optimization_example.py`)
+
+Demonstrates Tier 0/1/2 performance optimizations:
+
+```bash
+python optimization_example.py
+```
+
+## CLI Commands
+
+You can also profile directly from the command line:
+
+```bash
+# Profile a query
+praisonai profile query "What is 2+2?"
+
+# Profile with file grouping
+praisonai profile query "Hello" --show-files --limit 20
+
+# Profile imports
+praisonai profile imports
+
+# Profile startup time
+praisonai profile startup
+
+# Run comprehensive suite
+praisonai profile suite --quick
+
+# Create performance baseline
+praisonai profile snapshot --baseline
+
+# Compare against baseline
+praisonai profile snapshot current --compare
+
+# Show optimization status
+praisonai profile optimize --show
+```
+
+## Output
+
+- **Text output**: Human-readable timing breakdown
+- **JSON output**: Machine-readable for CI/CD (`--format json`)
+- **Artifacts**: Binary cProfile data and reports (`--save ./results`)
+- **Snapshots**: Performance baselines for regression detection
+
+## Performance Optimizations
+
+Enable opt-in optimizations via environment variables:
+
+```bash
+export PRAISONAI_LITE_MODE=1
+export PRAISONAI_SKIP_TYPE_VALIDATION=1
+export PRAISONAI_MINIMAL_IMPORTS=1
+```
@@ -0,0 +1,71 @@
+"""
+Basic Profiling Example for PraisonAI
+
+This example demonstrates how to use the unified profiler programmatically
+to measure agent performance.
+
+Run with:
+    python basic_profiling.py
+"""
+
+import os
+
+# Ensure API key is set
+if not os.environ.get("OPENAI_API_KEY"):
+    print("Please set OPENAI_API_KEY environment variable")
+    exit(1)
+
+from praisonai.cli.execution import (
+    ExecutionRequest,
+    Profiler,
+    ProfilerConfig,
+)
+
+
+def main():
+    """Run a profiled query and display results."""
+    
+    # Configure profiler (Layer 1 = basic profiling with function stats)
+    config = ProfilerConfig(
+        layer=1,             # 0=minimal, 1=basic, 2=deep
+        limit=15,            # Show top 15 functions
+        show_callers=False,  # Set True for caller info (layer 2)
+        show_callees=False,  # Set True for callee info (layer 2)
+    )
+    
+    # Create execution request
+    request = ExecutionRequest(
+        prompt="What is 2+2?",
+        agent_name="ProfiledAgent",
+    )
+    
+    # Run profiled query
+    print("🔬 Running profiled query...")
+    profiler = Profiler(config)
+    result, report = profiler.profile_sync(request)
+    
+    # Print formatted report
+    print("\n" + report.to_text())
+    
+    # Access timing data programmatically
+    print("\n📊 Timing Summary:")
+    print(f"  Imports:        {report.timing.imports_ms:>10.2f} ms")
+    print(f"  Agent Init:     {report.timing.agent_init_ms:>10.2f} ms")
+    print(f"  Execution:      {report.timing.execution_ms:>10.2f} ms")
+    print(f"  Total:          {report.timing.total_ms:>10.2f} ms")
+    
+    # Show top 5 functions
+    if report.functions:
+        print("\n🔥 Top 5 Functions by Cumulative Time:")
+        for i, func in enumerate(report.functions[:5], 1):
+            print(f"  {i}. {func.name}: {func.cumulative_time_ms:.2f}ms")
+    
+    # Show response
+    print(f"\n💬 Response: {result.output}")
+    
+    # Export to JSON
+    print("\n📄 JSON Export available via: report.to_json()")
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,223 @@
+"""
+Performance Optimization Example for PraisonAI
+
+This example demonstrates how to use the Tier 0/1/2 optimization features
+programmatically.
+
+Run with:
+    python optimization_example.py
+"""
+
+import os
+
+# Ensure API key is set
+if not os.environ.get("OPENAI_API_KEY"):
+    print("Please set OPENAI_API_KEY environment variable")
+    exit(1)
+
+
+def demo_provider_cache():
+    """Demonstrate provider caching (Tier 0)."""
+    print("\n" + "=" * 60)
+    print("Tier 0: Provider Cache Demo")
+    print("=" * 60)
+    
+    from praisonai.cli.features.profiler import (
+        get_provider_cache,
+        resolve_model_provider,
+    )
+    
+    cache = get_provider_cache()
+    cache.invalidate()  # Start fresh
+    
+    # Resolve some models
+    models = ['gpt-4o', 'claude-3-opus', 'gemini-pro', 'gpt-4o-mini']
+    
+    print("\nResolving models:")
+    for model in models:
+        provider, name = resolve_model_provider(model)
+        print(f"  {model} -> {provider}/{name}")
+    
+    # Check cache stats
+    stats = cache.get_stats()
+    print("\nCache Stats:")
+    print(f"  Hits: {stats['hits']}")
+    print(f"  Misses: {stats['misses']}")
+    print(f"  Size: {stats['size']}")
+
+
+def demo_lazy_importer():
+    """Demonstrate lazy imports (Tier 0)."""
+    print("\n" + "=" * 60)
+    print("Tier 0: Lazy Importer Demo")
+    print("=" * 60)
+    
+    from praisonai.cli.features.profiler import LazyImporter
+    import time
+    
+    LazyImporter.clear()
+    
+    # Time lazy import
+    start = time.perf_counter()
+    json_module = LazyImporter.get('json')
+    first_import = (time.perf_counter() - start) * 1000
+    
+    # Second import (cached)
+    start = time.perf_counter()
+    json_module2 = LazyImporter.get('json')
+    cached_import = (time.perf_counter() - start) * 1000
+    
+    print(f"\nFirst import: {first_import:.3f}ms")
+    print(f"Cached import: {cached_import:.3f}ms")
+    print(f"Same object: {json_module is json_module2}")
+
+
+def demo_client_pool():
+    """Demonstrate client pooling (Tier 1)."""
+    print("\n" + "=" * 60)
+    print("Tier 1: Client Pool Demo")
+    print("=" * 60)
+    
+    from praisonai.cli.features.profiler import ClientPool
+    
+    pool = ClientPool(max_size=3)
+    
+    # Create some mock clients
+    def create_client(name):
+        return {'name': name, 'created': True}
+    
+    # Get or create clients
+    client1 = pool.get_or_create('openai', lambda: create_client('openai'))
+    _ = pool.get_or_create('anthropic', lambda: create_client('anthropic'))
+    
+    # Reuse existing client
+    client1_again = pool.get_or_create('openai', lambda: create_client('openai_new'))
+    
+    print(f"\nPool size: {pool.size()}")
+    print(f"Client reused: {client1 is client1_again}")
+    print(f"Client1 name: {client1['name']}")
+
+
+def demo_prewarm():
+    """Demonstrate pre-warming (Tier 1)."""
+    print("\n" + "=" * 60)
+    print("Tier 1: Pre-warming Demo")
+    print("=" * 60)
+    
+    from praisonai.cli.features.profiler import PrewarmManager
+    import time
+    
+    PrewarmManager.reset()
+    
+    print(f"\nPre-warming enabled: {PrewarmManager.is_enabled()}")
+    
+    # Enable and prewarm
+    PrewarmManager.enable()
+    print(f"After enable: {PrewarmManager.is_enabled()}")
+    
+    # Prewarm OpenAI (runs in background)
+    PrewarmManager.prewarm_provider('openai')
+    print("Pre-warming OpenAI in background...")
+    
+    # Wait a bit for background thread
+    time.sleep(0.5)
+    
+    print(f"OpenAI prewarmed: {PrewarmManager.is_prewarmed('openai')}")
+    
+    PrewarmManager.reset()
+
+
+def demo_lite_mode():
+    """Demonstrate lite mode (Tier 2)."""
+    print("\n" + "=" * 60)
+    print("Tier 2: Lite Mode Demo")
+    print("=" * 60)
+    
+    from praisonai.cli.features.profiler import (
+        LiteModeConfig,
+        is_lite_mode,
+    )
+    
+    # Default config
+    config = LiteModeConfig()
+    print(f"\nDefault lite mode: {config.enabled}")
+    
+    # Config from env
+    print("\nTo enable lite mode, set environment variables:")
+    print("  export PRAISONAI_LITE_MODE=1")
+    print("  export PRAISONAI_SKIP_TYPE_VALIDATION=1")
+    
+    print(f"\nCurrent is_lite_mode(): {is_lite_mode()}")
+
+
+def demo_perf_snapshot():
+    """Demonstrate performance snapshots (Tier 2)."""
+    print("\n" + "=" * 60)
+    print("Tier 2: Performance Snapshot Demo")
+    print("=" * 60)
+    
+    from praisonai.cli.features.profiler import (
+        PerfSnapshot,
+        PerfComparison,
+        PerfSnapshotManager,
+    )
+    import tempfile
+    from pathlib import Path
+    
+    # Create manager with temp directory
+    with tempfile.TemporaryDirectory() as tmpdir:
+        manager = PerfSnapshotManager(snapshot_dir=Path(tmpdir))
+        
+        # Create baseline
+        baseline = PerfSnapshot(
+            timestamp='2025-01-01T00:00:00Z',
+            name='baseline',
+            startup_cold_ms=100.0,
+            startup_warm_ms=80.0,
+            import_time_ms=500.0,
+            query_time_ms=2000.0,
+            first_token_ms=300.0,
+        )
+        manager.save_baseline(baseline)
+        print("\nBaseline saved")
+        
+        # Create current snapshot (simulated improvement)
+        current = PerfSnapshot(
+            timestamp='2025-01-02T00:00:00Z',
+            name='current',
+            startup_cold_ms=95.0,  # 5% faster
+            startup_warm_ms=78.0,
+            import_time_ms=480.0,  # 4% faster
+            query_time_ms=1900.0,  # 5% faster
+            first_token_ms=280.0,
+        )
+        
+        # Compare
+        comparison = PerfComparison(baseline=baseline, current=current)
+        
+        print(f"\nStartup improvement: {-comparison.startup_cold_diff_pct:.1f}%")
+        print(f"Import improvement: {-comparison.import_time_diff_pct:.1f}%")
+        print(f"Query improvement: {-comparison.query_time_diff_pct:.1f}%")
+        print(f"Is regression: {comparison.is_regression()}")
+
+
+def main():
+    """Run all demos."""
+    print("=" * 60)
+    print("PraisonAI Performance Optimization Demo")
+    print("=" * 60)
+    
+    demo_provider_cache()
+    demo_lazy_importer()
+    demo_client_pool()
+    demo_prewarm()
+    demo_lite_mode()
+    demo_perf_snapshot()
+    
+    print("\n" + "=" * 60)
+    print("All demos complete!")
+    print("=" * 60)
+
+
+if __name__ == "__main__":
+    main()
@@ -0,0 +1,227 @@
+{
+  "timestamp": "2025-12-31T18:45:38.187539Z",
+  "metadata": {
+    "python_version": "3.12.11",
+    "platform": "macOS-15.7.4-arm64-arm-64bit",
+    "praisonai_version": "2.9.2",
+    "timestamp": "2025-12-31T18:45:38.226558Z"
+  },
+  "startup": {
+    "cold_ms": 80.9305000002496,
+    "warm_ms": 55.266541996388696
+  },
+  "import_analysis": [
+    {
+      "module": "praisonaiagents",
+      "self_us": 2419,
+      "cumulative_us": 1645390,
+      "self_ms": 2.419,
+      "cumulative_ms": 1645.39
+    },
+    {
+      "module": "praisonaiagents.workflows",
+      "self_us": 670,
+      "cumulative_us": 1375919,
+      "self_ms": 0.67,
+      "cumulative_ms": 1375.919
+    },
+    {
+      "module": "praisonaiagents.workflows.yaml_parser",
+      "self_us": 775,
+      "cumulative_us": 1369435,
+      "self_ms": 0.775,
+      "cumulative_ms": 1369.435
+    },
+    {
+      "module": "praisonaiagents.agent.agent",
+      "self_us": 19,
+      "cumulative_us": 1349174,
+      "self_ms": 0.019,
+      "cumulative_ms": 1349.174
+    },
+    {
+      "module": "praisonaiagents.agent",
+      "self_us": 537,
+      "cumulative_us": 1349155,
+      "self_ms": 0.537,
+      "cumulative_ms": 1349.155
+    },
+    {
+      "module": "praisonaiagents.agent.agent",
+      "self_us": 1453,
+      "cumulative_us": 1285299,
+      "self_ms": 1.453,
+      "cumulative_ms": 1285.299
+    },
+    {
+      "module": "praisonaiagents.llm.openai_client",
+      "self_us": 7887,
+      "cumulative_us": 1123578,
+      "self_ms": 7.887,
+      "cumulative_ms": 1123.578
+    },
+    {
+      "module": "openai",
+      "self_us": 1083,
+      "cumulative_us": 1115692,
+      "self_ms": 1.083,
+      "cumulative_ms": 1115.692
+    },
+    {
+      "module": "openai.types",
+      "self_us": 2606,
+      "cumulative_us": 957990,
+      "self_ms": 2.606,
+      "cumulative_ms": 957.99
+    },
+    {
+      "module": "openai.types.batch",
+      "self_us": 1787,
+      "cumulative_us": 485667,
+      "self_ms": 1.787,
+      "cumulative_ms": 485.667
+    },
+    {
+      "module": "openai._models",
+      "self_us": 13423,
+      "cumulative_us": 450798,
+      "self_ms": 13.423,
+      "cumulative_ms": 450.798
+    },
+    {
+      "module": "openai.types.eval_create_params",
+      "self_us": 644,
+      "cumulative_us": 393275,
+      "self_ms": 0.644,
+      "cumulative_ms": 393.275
+    },
+    {
+      "module": "openai.types.graders.grader_inputs_param",
+      "self_us": 17,
+      "cumulative_us": 359515,
+      "self_ms": 0.017,
+      "cumulative_ms": 359.515
+    },
+    {
+      "module": "openai.types.graders",
+      "self_us": 392,
+      "cumulative_us": 359498,
+      "self_ms": 0.392,
+      "cumulative_ms": 359.498
+    },
+    {
+      "module": "openai.types.graders.multi_grader",
+      "self_us": 3847,
+      "cumulative_us": 351251,
+      "self_ms": 3.847,
+      "cumulative_ms": 351.251
+    },
+    {
+      "module": "openai.types.graders.label_model_grader",
+      "self_us": 5450,
+      "cumulative_us": 337291,
+      "self_ms": 5.45,
+      "cumulative_ms": 337.291
+    },
+    {
+      "module": "openai.types.graders.grader_inputs",
+      "self_us": 545,
+      "cumulative_us": 331841,
+      "self_ms": 0.545,
+      "cumulative_ms": 331.841
+    },
+    {
+      "module": "openai.types.responses.response_input_text",
+      "self_us": 21,
+      "cumulative_us": 331297,
+      "self_ms": 0.021,
+      "cumulative_ms": 331.297
+    },
+    {
+      "module": "openai.types.responses",
+      "self_us": 1301,
+      "cumulative_us": 331277,
+      "self_ms": 1.301,
+      "cumulative_ms": 331.277
+    },
+    {
+      "module": "rich.logging",
+      "self_us": 1066,
+      "cumulative_us": 238256,
+      "self_ms": 1.066,
+      "cumulative_ms": 238.256
+    }
+  ],
+  "scenarios": [
+    {
+      "name": "simple_query",
+      "config": {
+        "prompt": "What is 2+2?",
+        "model": "default",
+        "stream": false,
+        "iterations": 2
+      },
+      "import_time_stats": {
+        "mean": 0.008833496394800022,
+        "min": 0.005916997906751931,
+        "max": 0.011749994882848114,
+        "stdev": 0.0041245517164382375,
+        "median": 0.008833496394800022
+      },
+      "total_time_stats": {
+        "mean": 1727.4090415012324,
+        "min": 1367.2264170018025,
+        "max": 2087.5916660006624,
+        "stdev": 509.3751524982296,
+        "median": 1727.4090415012324
+      },
+      "first_token_stats": null,
+      "raw_import_times_ms": [
+        0.011749994882848114,
+        0.005916997906751931
+      ],
+      "raw_total_times_ms": [
+        1367.2264170018025,
+        2087.5916660006624
+      ]
+    },
+    {
+      "name": "simple_stream",
+      "config": {
+        "prompt": "What is 2+2?",
+        "model": "default",
+        "stream": true,
+        "iterations": 2
+      },
+      "import_time_stats": {
+        "mean": 0.008792001608526334,
+        "min": 0.006375004886649549,
+        "max": 0.01120899833040312,
+        "stdev": 0.0034181495442894613,
+        "median": 0.008792001608526334
+      },
+      "total_time_stats": {
+        "mean": 2321.0982290001994,
+        "min": 1697.1089580038097,
+        "max": 2945.087499996589,
+        "stdev": 882.4540898183949,
+        "median": 2321.0982290001994
+      },
+      "first_token_stats": {
+        "mean": 2312.8511880022415,
+        "min": 1681.2440420035273,
+        "max": 2944.458334000956,
+        "stdev": 893.2273919631452,
+        "median": 2312.8511880022415
+      },
+      "raw_import_times_ms": [
+        0.01120899833040312,
+        0.006375004886649549
+      ],
+      "raw_total_times_ms": [
+        2945.087499996589,
+        1697.1089580038097
+      ]
+    }
+  ]
+}
@@ -0,0 +1,80 @@
+"""
+Profile Suite Example for PraisonAI
+
+This example demonstrates how to run a comprehensive profiling suite
+programmatically with multiple scenarios.
+
+Run with:
+    python suite_profiling.py
+"""
+
+import os
+
+# Ensure API key is set
+if not os.environ.get("OPENAI_API_KEY"):
+    print("Please set OPENAI_API_KEY environment variable")
+    exit(1)
+
+from praisonai.cli.features.profiler import (
+    ScenarioConfig,
+    run_profile_suite,
+)
+
+
+def main():
+    """Run a profiling suite with custom scenarios."""
+    
+    # Define custom scenarios
+    scenarios = [
+        ScenarioConfig(
+            name="simple_query",
+            prompt="What is 2+2?",
+            stream=False,
+            iterations=2,
+            warmup=1,
+        ),
+        ScenarioConfig(
+            name="simple_stream",
+            prompt="What is 2+2?",
+            stream=True,
+            iterations=2,
+            warmup=1,
+        ),
+    ]
+    
+    # Run suite with custom output directory
+    print("🔬 Running Profile Suite...")
+    result = run_profile_suite(
+        output_dir="./profile_results",
+        scenarios=scenarios,
+        verbose=True,
+    )
+    
+    # Print summary
+    print("\n" + "=" * 60)
+    print("Profile Suite Results")
+    print("=" * 60)
+    
+    print("\n📊 Startup Times:")
+    print(f"  Cold Start: {result.startup_cold_ms:.2f}ms")
+    print(f"  Warm Start: {result.startup_warm_ms:.2f}ms")
+    
+    if result.import_analysis:
+        print("\n📦 Top Import:")
+        print(f"  {result.import_analysis[0]['module']}: {result.import_analysis[0]['cumulative_ms']:.2f}ms")
+    
+    print("\n🎯 Scenario Results:")
+    for scenario in result.scenarios:
+        stats = scenario.get_stats(scenario.total_times)
+        print(f"  {scenario.name}:")
+        print(f"    Mean: {stats['mean']:.2f}ms")
+        print(f"    Min:  {stats['min']:.2f}ms")
+        print(f"    Max:  {stats['max']:.2f}ms")
+    
+    print("\n✅ Results saved to: ./profile_results/")
+    print("   - suite_results.json (machine-readable)")
+    print("   - suite_report.txt (human-readable)")
+
+
+if __name__ == "__main__":
+    main()
@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"

 [project]
 name = "praisonaiagents"
-version = "0.5.3"
+version = "0.6.0"
 description = "Praison AI agents for completing complex tasks with Self Reflection Agents"
 readme = "README.md"
 requires-python = ">=3.10"
@@ -2422,7 +2422,7 @@ wheels = [

 [[package]]
 name = "praisonaiagents"
-version = "0.5.3"
+version = "0.6.0"
 source = { editable = "." }
 dependencies = [
    { name = "aiohttp" },
@@ -3,8 +3,8 @@ class Praisonai < Formula
  
    desc "AI tools for various AI applications"
    homepage "https://github.com/MervinPraison/PraisonAI"
-    url "https://github.com/MervinPraison/PraisonAI/archive/refs/tags/v2.9.2.tar.gz"
-    sha256 `curl -sL https://github.com/MervinPraison/PraisonAI/archive/refs/tags/v2.9.2.tar.gz | shasum -a 256`.split.first
+    url "https://github.com/MervinPraison/PraisonAI/archive/refs/tags/v3.0.2.tar.gz"
+    sha256 `curl -sL https://github.com/MervinPraison/PraisonAI/archive/refs/tags/v3.0.2.tar.gz | shasum -a 256`.split.first
    license "MIT"
  
    depends_on "[email protected]"
@@ -176,6 +176,7 @@ def register_commands():
    from .commands.serve import app as serve_app
    from .commands.schedule import app as schedule_app
    from .commands.run import app as run_app
+    from .commands.profile import app as profile_app
    
    # Import TUI and queue commands
    from .features.tui.debug import create_debug_app as create_tui_debug_app
@@ -197,6 +198,7 @@ def register_commands():
    app.add_typer(serve_app, name="serve", help="API server management")
    app.add_typer(schedule_app, name="schedule", help="Scheduler management")
    app.add_typer(run_app, name="run", help="Run agents")
+    app.add_typer(profile_app, name="profile", help="Performance profiling and diagnostics")
    
    # Register TUI and queue commands
    tui_app = create_tui_debug_app()
@@ -0,0 +1,466 @@
+"""
+Profile command group for PraisonAI CLI.
+
+Provides detailed cProfile-based profiling for query execution.
+"""
+
+import sys
+from typing import Optional
+
+import typer
+
+app = typer.Typer(help="Performance profiling and diagnostics")
+
+
+@app.command("query")
+def profile_query(
+    prompt: str = typer.Argument(..., help="Prompt to profile"),
+    model: Optional[str] = typer.Option(None, "--model", "-m", help="Model to use"),
+    stream: bool = typer.Option(False, "--stream/--no-stream", help="Use streaming mode"),
+    deep: bool = typer.Option(False, "--deep", help="Enable deep call tracing (higher overhead)"),
+    limit: int = typer.Option(30, "--limit", "-n", help="Top N functions to show"),
+    sort: str = typer.Option("cumulative", "--sort", "-s", help="Sort by: cumulative or tottime"),
+    show_callers: bool = typer.Option(False, "--show-callers", help="Show caller functions"),
+    show_callees: bool = typer.Option(False, "--show-callees", help="Show callee functions"),
+    save: Optional[str] = typer.Option(None, "--save", help="Save artifacts to path (creates .prof, .txt, .json)"),
+    output_format: str = typer.Option("text", "--format", "-f", help="Output format: text or json"),
+):
+    """
+    Profile a query execution with detailed timing breakdown.
+    
+    Uses the unified profiling architecture for consistent results across
+    CLI direct invocation (--profile) and this command.
+    
+    Shows per-function timing, call graphs, and latency metrics.
+    
+    Examples:
+        praisonai profile query "What is 2+2?"
+        praisonai profile query "Hello" --limit 20
+        praisonai profile query "Test" --deep --show-callers --show-callees
+        praisonai profile query "Test" --save ./profile_results
+    """
+    # Use unified profiler for consistent results
+    try:
+        from ..execution import ExecutionRequest, Profiler, ProfilerConfig as UnifiedProfilerConfig
+    except ImportError as e:
+        typer.echo(f"Error: Unified profiler not available: {e}", err=True)
+        raise typer.Exit(1)
+    
+    # Warn about deep tracing overhead
+    if deep:
+        typer.echo("⚠️  Deep call tracing enabled - this adds significant overhead", err=True)
+    
+    # Create unified execution request
+    request = ExecutionRequest(
+        prompt=prompt,
+        agent_name="ProfiledAgent",
+        model=model,
+        stream=stream,
+    )
+    
+    # Create unified profiler config
+    config = UnifiedProfilerConfig(
+        layer=2 if deep else 1,
+        limit=limit,
+        sort_by=sort,
+        show_callers=show_callers,
+        show_callees=show_callees,
+        output_format=output_format,
+        save_path=save,
+    )
+    
+    # Run profiler
+    typer.echo("🔬 Starting profiled execution...", err=True)
+    
+    try:
+        profiler = Profiler(config)
+        result, report = profiler.profile_sync(request, invocation_method="profile_command")
+    except ImportError as e:
+        typer.echo(f"Error: {e}", err=True)
+        typer.echo("Install praisonaiagents: pip install praisonaiagents", err=True)
+        raise typer.Exit(1)
+    except Exception as e:
+        typer.echo(f"Error during profiling: {e}", err=True)
+        raise typer.Exit(1)
+    
+    # Output results
+    if output_format == "json":
+        typer.echo(report.to_json())
+    else:
+        typer.echo(report.to_text())
+    
+    # Save artifacts if requested
+    if save:
+        try:
+            import os
+            os.makedirs(save, exist_ok=True)
+            
+            # Save JSON report
+            json_path = os.path.join(save, "profile_report.json")
+            with open(json_path, 'w') as f:
+                f.write(report.to_json())
+            
+            # Save text report
+            txt_path = os.path.join(save, "profile_report.txt")
+            with open(txt_path, 'w') as f:
+                f.write(report.to_text())
+            
+            typer.echo("\n✅ Artifacts saved:", err=True)
+            typer.echo(f"   {json_path}", err=True)
+            typer.echo(f"   {txt_path}", err=True)
+        except Exception as e:
+            typer.echo(f"Warning: Failed to save artifacts: {e}", err=True)
+
+
+@app.command("imports")
+def profile_imports():
+    """
+    Profile module import times.
+    
+    Shows which modules take longest to import, useful for optimizing startup time.
+    """
+    import subprocess
+    import re
+    
+    typer.echo("🔬 Profiling import times...", err=True)
+    
+    try:
+        result = subprocess.run(
+            [sys.executable, "-X", "importtime", "-c", "import praisonaiagents"],
+            capture_output=True,
+            text=True,
+            timeout=60,
+        )
+        
+        # Parse import times
+        import_times = []
+        for line in result.stderr.split('\n'):
+            if 'import time:' in line:
+                match = re.search(r'import time:\s+(\d+)\s+\|\s+(\d+)\s+\|\s+(.+)', line)
+                if match:
+                    self_time = int(match.group(1))
+                    cumulative = int(match.group(2))
+                    module = match.group(3).strip()
+                    import_times.append((module, self_time, cumulative))
+        
+        # Sort by cumulative time
+        import_times.sort(key=lambda x: x[2], reverse=True)
+        
+        # Display
+        typer.echo("\n" + "=" * 70)
+        typer.echo("Import Time Analysis")
+        typer.echo("=" * 70)
+        typer.echo(f"{'Module':<45} {'Self (μs)':>10} {'Cumul (μs)':>12}")
+        typer.echo("-" * 70)
+        
+        for module, self_time, cumulative in import_times[:30]:
+            module_name = module
+            if len(module_name) > 43:
+                module_name = "..." + module_name[-40:]
+            typer.echo(f"{module_name:<45} {self_time:>10} {cumulative:>12}")
+        
+        typer.echo("-" * 70)
+        
+        if import_times:
+            total_time = max(t[2] for t in import_times)
+            typer.echo(f"Total import time: {total_time / 1000:.2f} ms")
+        
+    except subprocess.TimeoutExpired:
+        typer.echo("Error: Import profiling timed out", err=True)
+        raise typer.Exit(1)
+    except Exception as e:
+        typer.echo(f"Error: {e}", err=True)
+        raise typer.Exit(1)
+
+
+@app.command("startup")
+def profile_startup():
+    """
+    Profile CLI startup time.
+    
+    Measures time from CLI invocation to ready state.
+    """
+    import time
+    import subprocess
+    
+    typer.echo("🔬 Profiling startup time...", err=True)
+    
+    # Measure cold start (new process)
+    start = time.perf_counter()
+    subprocess.run(
+        [sys.executable, "-c", "import praisonai; import praisonai.cli"],
+        capture_output=True,
+        text=True,
+    )
+    cold_start = (time.perf_counter() - start) * 1000
+    
+    # Measure warm start (imports cached)
+    start = time.perf_counter()
+    subprocess.run(
+        [sys.executable, "-c", "import praisonai; import praisonai.cli"],
+        capture_output=True,
+        text=True,
+    )
+    warm_start = (time.perf_counter() - start) * 1000
+    
+    typer.echo("\n" + "=" * 50)
+    typer.echo("Startup Time Analysis")
+    typer.echo("=" * 50)
+    typer.echo(f"Cold Start:  {cold_start:>10.2f} ms")
+    typer.echo(f"Warm Start:  {warm_start:>10.2f} ms")
+    typer.echo("=" * 50)
+    
+    if cold_start > 1000:
+        typer.echo("\n⚠️  Cold start > 1s - consider lazy imports", err=True)
+
+
+@app.command("suite")
+def profile_suite(
+    output_dir: Optional[str] = typer.Option("/tmp/praisonai_profile_suite", "--output", "-o", help="Output directory for results"),
+    iterations: int = typer.Option(3, "--iterations", "-n", help="Iterations per scenario"),
+    quick: bool = typer.Option(False, "--quick", help="Quick mode (fewer iterations)"),
+):
+    """
+    Run a comprehensive profiling suite.
+    
+    Runs multiple scenarios (streaming/non-streaming, simple/complex prompts)
+    and produces aggregated statistics and reports.
+    
+    Examples:
+        praisonai profile suite
+        praisonai profile suite --output ./my_profile_results
+        praisonai profile suite --quick
+    """
+    try:
+        from ..features.profiler import run_profile_suite, ScenarioConfig
+    except ImportError as e:
+        typer.echo(f"Error: Profiler module not available: {e}", err=True)
+        raise typer.Exit(1)
+    
+    # Define scenarios
+    iters = 1 if quick else iterations
+    scenarios = [
+        ScenarioConfig(name="simple_non_stream", prompt="hi", stream=False, iterations=iters, warmup=0 if quick else 1),
+        ScenarioConfig(name="simple_stream", prompt="hi", stream=True, iterations=iters, warmup=0 if quick else 1),
+    ]
+    
+    if not quick:
+        scenarios.extend([
+            ScenarioConfig(name="medium_non_stream", prompt="Explain Python in 2 sentences", stream=False, iterations=max(1, iters-1)),
+            ScenarioConfig(name="medium_stream", prompt="Explain Python in 2 sentences", stream=True, iterations=max(1, iters-1)),
+        ])
+    
+    typer.echo("🔬 Running Profile Suite...", err=True)
+    typer.echo(f"   Output: {output_dir}", err=True)
+    typer.echo(f"   Scenarios: {len(scenarios)}", err=True)
+    typer.echo(f"   Iterations: {iterations}", err=True)
+    
+    try:
+        result = run_profile_suite(
+            output_dir=output_dir,
+            scenarios=scenarios,
+            verbose=True,
+        )
+        
+        # Print summary
+        typer.echo("\n" + "=" * 60)
+        typer.echo("Profile Suite Summary")
+        typer.echo("=" * 60)
+        typer.echo(f"Startup Cold: {result.startup_cold_ms:.2f}ms")
+        typer.echo(f"Startup Warm: {result.startup_warm_ms:.2f}ms")
+        
+        if result.import_analysis:
+            typer.echo(f"\nTop Import: {result.import_analysis[0]['module']}")
+            typer.echo(f"  Time: {result.import_analysis[0]['cumulative_ms']:.2f}ms")
+        
+        typer.echo("\nScenario Results:")
+        for scenario in result.scenarios:
+            stats = scenario.get_stats(scenario.total_times)
+            typer.echo(f"  {scenario.name}: {stats['mean']:.2f}ms (±{stats['stdev']:.2f}ms)")
+        
+        typer.echo(f"\n✅ Full results saved to: {output_dir}")
+        
+    except Exception as e:
+        typer.echo(f"Error running suite: {e}", err=True)
+        raise typer.Exit(1)
+
+
+@app.command("snapshot")
+def profile_snapshot(
+    name: str = typer.Argument("current", help="Name for the snapshot"),
+    baseline: bool = typer.Option(False, "--baseline", "-b", help="Save as baseline for comparison"),
+    compare: bool = typer.Option(False, "--compare", "-c", help="Compare against baseline"),
+    output_format: str = typer.Option("text", "--format", "-f", help="Output format: text or json"),
+):
+    """
+    Create or compare performance snapshots.
+    
+    Snapshots record timing data for baseline comparison.
+    Use --baseline to save a reference point, then --compare to check for regressions.
+    
+    Examples:
+        praisonai profile snapshot --baseline
+        praisonai profile snapshot current --compare
+        praisonai profile snapshot v2.0 --format json
+    """
+    try:
+        from ..features.profiler import (
+            run_profile_suite,
+            ScenarioConfig,
+            PerfSnapshotManager,
+            create_snapshot_from_suite,
+            format_comparison_report,
+        )
+    except ImportError as e:
+        typer.echo(f"Error: Profiler module not available: {e}", err=True)
+        raise typer.Exit(1)
+    
+    manager = PerfSnapshotManager()
+    
+    if compare:
+        # Load baseline and compare
+        baseline_snap = manager.load_baseline()
+        if not baseline_snap:
+            typer.echo("❌ No baseline found. Create one with: praisonai profile snapshot --baseline", err=True)
+            raise typer.Exit(1)
+        
+        typer.echo("🔬 Running current profile for comparison...", err=True)
+        
+        # Run quick suite
+        scenarios = [
+            ScenarioConfig(name="simple", prompt="hi", stream=False, iterations=2, warmup=1),
+        ]
+        result = run_profile_suite(
+            output_dir="/tmp/praisonai_snapshot_compare",
+            scenarios=scenarios,
+            verbose=False,
+        )
+        
+        current_snap = create_snapshot_from_suite(result, name)
+        comparison = manager.compare(current_snap)
+        
+        if output_format == "json":
+            import json
+            typer.echo(json.dumps(comparison.to_dict(), indent=2))
+        else:
+            typer.echo(format_comparison_report(comparison))
+        
+        if comparison.is_regression():
+            raise typer.Exit(1)  # Exit with error on regression
+    else:
+        # Create new snapshot
+        typer.echo(f"🔬 Creating performance snapshot '{name}'...", err=True)
+        
+        scenarios = [
+            ScenarioConfig(name="simple", prompt="hi", stream=False, iterations=3, warmup=1),
+            ScenarioConfig(name="stream", prompt="hi", stream=True, iterations=2, warmup=1),
+        ]
+        result = run_profile_suite(
+            output_dir="/tmp/praisonai_snapshot",
+            scenarios=scenarios,
+            verbose=True,
+        )
+        
+        snapshot = create_snapshot_from_suite(result, name)
+        
+        if baseline:
+            path = manager.save_baseline(snapshot)
+            typer.echo(f"\n✅ Baseline saved to: {path}")
+        else:
+            path = manager.save(snapshot)
+            typer.echo(f"\n✅ Snapshot saved to: {path}")
+        
+        if output_format == "json":
+            import json
+            typer.echo(json.dumps(snapshot.to_dict(), indent=2))
+        else:
+            typer.echo("\nSnapshot Summary:")
+            typer.echo(f"  Startup Cold: {snapshot.startup_cold_ms:.2f}ms")
+            typer.echo(f"  Import Time:  {snapshot.import_time_ms:.2f}ms")
+            typer.echo(f"  Query Time:   {snapshot.query_time_ms:.2f}ms")
+
+
+@app.command("optimize")
+def profile_optimize(
+    prewarm: bool = typer.Option(False, "--prewarm", help="Enable provider pre-warming"),
+    lite: bool = typer.Option(False, "--lite", help="Enable lite mode (skip type validation)"),
+    show_config: bool = typer.Option(False, "--show", help="Show current optimization config"),
+):
+    """
+    Configure performance optimizations.
+    
+    All optimizations are opt-in and safe by default.
+    
+    Examples:
+        praisonai profile optimize --show
+        praisonai profile optimize --prewarm
+        praisonai profile optimize --lite
+    """
+    try:
+        from ..features.profiler import (
+            PrewarmManager,
+            get_lite_mode_config,
+            get_provider_cache,
+        )
+    except ImportError as e:
+        typer.echo(f"Error: Profiler module not available: {e}", err=True)
+        raise typer.Exit(1)
+    
+    if show_config:
+        typer.echo("=" * 50)
+        typer.echo("Performance Optimization Status")
+        typer.echo("=" * 50)
+        
+        lite_config = get_lite_mode_config()
+        typer.echo(f"\nLite Mode: {'Enabled' if lite_config.enabled else 'Disabled'}")
+        if lite_config.enabled:
+            typer.echo(f"  Skip Type Validation: {lite_config.skip_type_validation}")
+            typer.echo(f"  Skip Model Validation: {lite_config.skip_model_validation}")
+            typer.echo(f"  Minimal Imports: {lite_config.minimal_imports}")
+        
+        typer.echo(f"\nPre-warming: {'Enabled' if PrewarmManager.is_enabled() else 'Disabled'}")
+        
+        cache = get_provider_cache()
+        stats = cache.get_stats()
+        typer.echo("\nProvider Cache:")
+        typer.echo(f"  Hits: {stats['hits']}")
+        typer.echo(f"  Misses: {stats['misses']}")
+        typer.echo(f"  Size: {stats['size']}")
+        
+        typer.echo("\n" + "=" * 50)
+        typer.echo("\nTo enable optimizations, set environment variables:")
+        typer.echo("  PRAISONAI_LITE_MODE=1")
+        typer.echo("  PRAISONAI_SKIP_TYPE_VALIDATION=1")
+        typer.echo("  PRAISONAI_MINIMAL_IMPORTS=1")
+        return
+    
+    if prewarm:
+        typer.echo("🔥 Enabling provider pre-warming...")
+        PrewarmManager.enable()
+        
+        import os
+        if os.environ.get('OPENAI_API_KEY'):
+            typer.echo("   Pre-warming OpenAI...")
+            PrewarmManager.prewarm_provider('openai')
+        if os.environ.get('ANTHROPIC_API_KEY'):
+            typer.echo("   Pre-warming Anthropic...")
+            PrewarmManager.prewarm_provider('anthropic')
+        
+        typer.echo("✅ Pre-warming initiated (runs in background)")
+    
+    if lite:
+        typer.echo("⚡ Lite mode configuration:")
+        typer.echo("   Set these environment variables to enable:")
+        typer.echo("   export PRAISONAI_LITE_MODE=1")
+        typer.echo("   export PRAISONAI_SKIP_TYPE_VALIDATION=1")
+        typer.echo("   export PRAISONAI_MINIMAL_IMPORTS=1")
+    
+    if not prewarm and not lite and not show_config:
+        typer.echo("Use --show to see current config, or --prewarm/--lite to enable optimizations")
+
+
+@app.callback(invoke_without_command=True)
+def profile_callback(ctx: typer.Context):
+    """Show profile help if no subcommand."""
+    if ctx.invoked_subcommand is None:
+        typer.echo(ctx.get_help())
@@ -0,0 +1,99 @@
+"""
+PraisonAI Unified Execution Module.
+
+Provides a single, unified execution primitive for all CLI modes:
+- CLI direct prompt
+- Profile command
+- TUI workers
+
+Design Principles:
+1. Profiling observes, never alters execution
+2. Zero overhead when profiling disabled
+3. Sync and async are first-class citizens
+4. Fail-safe, not fail-fast
+5. Bounded resource usage
+6. Schema stability with versioning
+"""
+
+from typing import TYPE_CHECKING
+
+if TYPE_CHECKING:
+    from .request import ExecutionRequest
+    from .result import ExecutionResult
+    from .core import execute_sync, execute_async
+    from .profiler import Profiler, ProfilerConfig, ProfileReport
+
+# Lazy loading to maintain zero-overhead when not used
+_lazy_cache = {}
+
+
+def __getattr__(name: str):
+    """Lazy load execution components."""
+    global _lazy_cache
+    
+    if name in _lazy_cache:
+        return _lazy_cache[name]
+    
+    if name == "ExecutionRequest":
+        from .request import ExecutionRequest
+        _lazy_cache[name] = ExecutionRequest
+        return ExecutionRequest
+    elif name == "ExecutionResult":
+        from .result import ExecutionResult
+        _lazy_cache[name] = ExecutionResult
+        return ExecutionResult
+    elif name == "execute_sync":
+        from .core import execute_sync
+        _lazy_cache[name] = execute_sync
+        return execute_sync
+    elif name == "execute_async":
+        from .core import execute_async
+        _lazy_cache[name] = execute_async
+        return execute_async
+    elif name == "Profiler":
+        from .profiler import Profiler
+        _lazy_cache[name] = Profiler
+        return Profiler
+    elif name == "ProfilerConfig":
+        from .profiler import ProfilerConfig
+        _lazy_cache[name] = ProfilerConfig
+        return ProfilerConfig
+    elif name == "ProfileReport":
+        from .profiler import ProfileReport
+        _lazy_cache[name] = ProfileReport
+        return ProfileReport
+    elif name == "TimingBreakdown":
+        from .profiler import TimingBreakdown
+        _lazy_cache[name] = TimingBreakdown
+        return TimingBreakdown
+    elif name == "InvocationInfo":
+        from .profiler import InvocationInfo
+        _lazy_cache[name] = InvocationInfo
+        return InvocationInfo
+    elif name == "DecisionTrace":
+        from .profiler import DecisionTrace
+        _lazy_cache[name] = DecisionTrace
+        return DecisionTrace
+    elif name == "ModuleBreakdown":
+        from .profiler import ModuleBreakdown
+        _lazy_cache[name] = ModuleBreakdown
+        return ModuleBreakdown
+    
+    raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
+
+
+__all__ = [
+    # Core execution
+    "ExecutionRequest",
+    "ExecutionResult",
+    "execute_sync",
+    "execute_async",
+    # Profiling
+    "Profiler",
+    "ProfilerConfig",
+    "ProfileReport",
+    "TimingBreakdown",
+    "InvocationInfo",
+    "DecisionTrace",
+    "ModuleBreakdown",
+]
@@ -0,0 +1,210 @@
+"""
+Core Execution Module.
+
+Provides the unified execution primitives:
+- _execute_core(): Single implementation shared by all paths
+- execute_sync(): Synchronous wrapper for CLI
+- execute_async(): Asynchronous wrapper for TUI workers
+"""
+
+import asyncio
+import time
+import uuid
+from typing import Optional, AsyncGenerator
+
+from .request import ExecutionRequest
+from .result import ExecutionResult
+
+
+def _generate_run_id() -> str:
+    """Generate a unique run ID."""
+    return str(uuid.uuid4())[:8]
+
+
+def _execute_core(
+    request: ExecutionRequest,
+    timing: Optional[dict] = None,
+) -> ExecutionResult:
+    """
+    THE SINGLE EXECUTION IMPLEMENTATION.
+    
+    All execution modes call this function.
+    Profiler wraps this function, never injects into it.
+    
+    Args:
+        request: The execution request
+        timing: Optional dict to record timing phases (for profiling)
+    
+    Returns:
+        ExecutionResult with output and metadata
+    """
+    run_id = _generate_run_id()
+    start_time = time.perf_counter()
+    
+    if timing is not None:
+        timing["core_start"] = start_time
+    
+    try:
+        # Phase 1: Import (measured inline)
+        import_start = time.perf_counter()
+        from praisonaiagents import Agent
+        import_end = time.perf_counter()
+        
+        if timing is not None:
+            timing["imports_ms"] = (import_end - import_start) * 1000
+        
+        # Phase 2: Agent Construction
+        agent_start = time.perf_counter()
+        
+        agent_config = {
+            "name": request.agent_name,
+            "verbose": False,
+        }
+        
+        if request.agent_instructions:
+            agent_config["instructions"] = request.agent_instructions
+        
+        if request.model:
+            agent_config["llm"] = request.model
+        
+        if request.tools:
+            agent_config["tools"] = list(request.tools)
+        
+        agent = Agent(**agent_config)
+        agent_end = time.perf_counter()
+        
+        if timing is not None:
+            timing["agent_init_ms"] = (agent_end - agent_start) * 1000
+        
+        # Phase 3: Execution
+        exec_start = time.perf_counter()
+        
+        if request.stream:
+            # For streaming, collect all chunks
+            chunks = []
+            if hasattr(agent, '_start_stream'):
+                for chunk in agent._start_stream(request.prompt):
+                    chunks.append(chunk)
+            else:
+                # Fallback to non-streaming
+                result = agent.start(request.prompt)
+                chunks = [result] if result else []
+            output = "".join(chunks)
+        else:
+            output = agent.start(request.prompt)
+        
+        exec_end = time.perf_counter()
+        
+        if timing is not None:
+            timing["execution_ms"] = (exec_end - exec_start) * 1000
+        
+        end_time = time.perf_counter()
+        
+        if timing is not None:
+            timing["total_ms"] = (end_time - start_time) * 1000
+        
+        return ExecutionResult(
+            output=output or "",
+            run_id=run_id,
+            success=True,
+            start_time=start_time,
+            end_time=end_time,
+            metadata={
+                "model": request.model,
+                "stream": request.stream,
+            },
+        )
+        
+    except Exception as e:
+        end_time = time.perf_counter()
+        
+        if timing is not None:
+            timing["total_ms"] = (end_time - start_time) * 1000
+            timing["error"] = str(e)
+        
+        return ExecutionResult.from_error(
+            error=str(e),
+            run_id=run_id,
+        )
+
+
+def execute_sync(request: ExecutionRequest) -> ExecutionResult:
+    """
+    Synchronous execution wrapper.
+    
+    Used by:
+    - CLI direct prompt
+    - Profile command
+    
+    Args:
+        request: The execution request
+    
+    Returns:
+        ExecutionResult with output and metadata
+    """
+    return _execute_core(request)
+
+
+async def execute_async(request: ExecutionRequest) -> ExecutionResult:
+    """
+    Asynchronous execution wrapper.
+    
+    Used by:
+    - TUI workers
+    - Queue system
+    
+    Args:
+        request: The execution request
+    
+    Returns:
+        ExecutionResult with output and metadata
+    """
+    # Run the synchronous core in a thread pool to avoid blocking
+    loop = asyncio.get_event_loop()
+    return await loop.run_in_executor(None, _execute_core, request, None)
+
+
+async def execute_async_stream(
+    request: ExecutionRequest,
+) -> AsyncGenerator[str, None]:
+    """
+    Asynchronous streaming execution.
+    
+    Yields chunks as they arrive.
+    
+    Args:
+        request: The execution request (stream=True recommended)
+    
+    Yields:
+        String chunks of the response
+    """
+    try:
+        from praisonaiagents import Agent
+        
+        agent_config = {
+            "name": request.agent_name,
+            "verbose": False,
+        }
+        
+        if request.agent_instructions:
+            agent_config["instructions"] = request.agent_instructions
+        
+        if request.model:
+            agent_config["llm"] = request.model
+        
+        if request.tools:
+            agent_config["tools"] = list(request.tools)
+        
+        agent = Agent(**agent_config)
+        
+        if hasattr(agent, '_start_stream'):
+            for chunk in agent._start_stream(request.prompt):
+                yield chunk
+                await asyncio.sleep(0)  # Yield control
+        else:
+            result = agent.start(request.prompt)
+            if result:
+                yield result
+                
+    except Exception as e:
+        yield f"Error: {e}"
@@ -0,0 +1,814 @@
+"""
+Profiler Module.
+
+Provides external profiling that wraps execution.
+Execution code has ZERO awareness of profiling.
+
+Design Principles:
+1. Profiler wraps execution, does not inject into it
+2. Layered profiling with explicit overhead
+3. Bounded resource usage
+4. Versioned, stable output schema
+"""
+
+import cProfile
+import pstats
+import time
+import platform
+from dataclasses import dataclass, field
+from datetime import datetime
+from typing import Any, Dict, List, Optional, Tuple
+
+from .request import ExecutionRequest
+from .result import ExecutionResult
+from .core import _execute_core
+
+
+# Schema version - MUST be incremented for breaking changes
+SCHEMA_VERSION = "1.0"
+
+# Bounds for profile data
+MAX_FUNCTION_STATS = 1000
+MAX_CALL_GRAPH_EDGES = 5000
+MAX_NETWORK_REQUESTS = 100
+MAX_IMPORT_MODULES = 200
+
+
+@dataclass
+class ProfilerConfig:
+    """
+    Profiler configuration with explicit layer control.
+    
+    Layers:
+    - 0: Wall-clock phases only (<1ms overhead)
+    - 1: + cProfile function stats (~5% overhead)
+    - 2: + Call graph callers/callees (~15% overhead)
+    
+    Network timing is orthogonal (opt-in via track_network).
+    """
+    
+    # Layer control (0, 1, or 2)
+    layer: int = 1
+    
+    # Layer 1+ options
+    limit: int = 30  # Max functions to show
+    sort_by: str = "cumulative"  # cumulative, time, calls
+    
+    # Layer 2 options
+    show_callers: bool = False
+    show_callees: bool = False
+    
+    # Network timing (orthogonal to layers)
+    track_network: bool = False
+    
+    # Output options
+    output_format: str = "text"  # text, json
+    save_path: Optional[str] = None
+    
+    @classmethod
+    def from_flags(
+        cls,
+        profile: bool = False,
+        deep: bool = False,
+        network: bool = False,
+    ) -> 'ProfilerConfig':
+        """Create config from CLI flags."""
+        if deep:
+            return cls(
+                layer=2,
+                show_callers=True,
+                show_callees=True,
+                track_network=network,
+            )
+        elif profile:
+            return cls(layer=1, track_network=network)
+        else:
+            return cls(layer=0, track_network=network)
+
+
+@dataclass
+class TimingBreakdown:
+    """
+    Phase timing breakdown with full timeline visibility.
+    
+    All fields in milliseconds.
+    Timeline order: cli_entry -> cli_parse -> routing -> imports -> agent_init -> 
+                    network_start -> first_token -> first_output -> execution_end
+    """
+    
+    # MANDATORY fields (always present)
+    total_ms: float = 0.0
+    imports_ms: float = 0.0
+    agent_init_ms: float = 0.0
+    execution_ms: float = 0.0
+    
+    # TIMELINE PHASES (new - for full visibility)
+    cli_entry_ms: float = 0.0      # Time from ENTER to CLI entry point
+    cli_parse_ms: float = 0.0      # Argparse/Typer routing time
+    routing_ms: float = 0.0        # Command routing time
+    network_start_ms: float = 0.0  # When network request started (relative to start)
+    first_token_ms: float = 0.0    # Time to first token (streaming)
+    first_output_ms: float = 0.0   # Time to First Response (TTFR) - first visible output
+    
+    # OPTIONAL fields (may be 0)
+    tool_time_ms: float = 0.0      # If tools were called
+    
+    @property
+    def time_to_first_response_ms(self) -> float:
+        """Time to First Response (TTFR) - the key user-perceived latency metric."""
+        if self.first_output_ms > 0:
+            return self.first_output_ms
+        # Fallback: use execution end if first_output not captured
+        return self.total_ms
+    
+    def to_dict(self) -> Dict[str, float]:
+        """Convert to dictionary."""
+        return {
+            "total_ms": self.total_ms,
+            "imports_ms": self.imports_ms,
+            "agent_init_ms": self.agent_init_ms,
+            "execution_ms": self.execution_ms,
+            "cli_entry_ms": self.cli_entry_ms,
+            "cli_parse_ms": self.cli_parse_ms,
+            "routing_ms": self.routing_ms,
+            "network_start_ms": self.network_start_ms,
+            "first_token_ms": self.first_token_ms,
+            "first_output_ms": self.first_output_ms,
+            "time_to_first_response_ms": self.time_to_first_response_ms,
+            "tool_time_ms": self.tool_time_ms,
+        }
+    
+    def to_timeline(self) -> List[Tuple[str, float, float]]:
+        """
+        Get ordered timeline of phases.
+        
+        Returns list of (phase_name, start_ms, duration_ms) tuples.
+        """
+        timeline = []
+        cursor = 0.0
+        
+        # CLI Entry (if captured)
+        if self.cli_entry_ms > 0:
+            timeline.append(("CLI Entry", 0.0, self.cli_entry_ms))
+            cursor = self.cli_entry_ms
+        
+        # CLI Parse
+        if self.cli_parse_ms > 0:
+            timeline.append(("CLI Parse", cursor, self.cli_parse_ms))
+            cursor += self.cli_parse_ms
+        
+        # Routing
+        if self.routing_ms > 0:
+            timeline.append(("Routing", cursor, self.routing_ms))
+            cursor += self.routing_ms
+        
+        # Imports
+        if self.imports_ms > 0:
+            timeline.append(("Imports", cursor, self.imports_ms))
+            cursor += self.imports_ms
+        
+        # Agent Init
+        if self.agent_init_ms > 0:
+            timeline.append(("Agent Init", cursor, self.agent_init_ms))
+            cursor += self.agent_init_ms
+        
+        # Network to First Token (if streaming)
+        if self.network_start_ms > 0 and self.first_token_ms > 0:
+            network_duration = self.first_token_ms - self.network_start_ms
+            if network_duration > 0:
+                timeline.append(("Network → First Token", self.network_start_ms, network_duration))
+        
+        # Execution
+        if self.execution_ms > 0:
+            timeline.append(("Execution", cursor, self.execution_ms))
+        
+        return timeline
+
+
+@dataclass
+class FunctionStat:
+    """Statistics for a single function."""
+    
+    name: str
+    file: str
+    line: int
+    calls: int
+    total_time_ms: float
+    cumulative_time_ms: float
+    
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary."""
+        return {
+            "name": self.name,
+            "file": self.file,
+            "line": self.line,
+            "calls": self.calls,
+            "total_time_ms": self.total_time_ms,
+            "cumulative_time_ms": self.cumulative_time_ms,
+        }
+
+
+@dataclass
+class CallGraph:
+    """Call graph data (callers and callees)."""
+    
+    callers: Dict[str, List[str]] = field(default_factory=dict)
+    callees: Dict[str, List[str]] = field(default_factory=dict)
+    
+    @property
+    def edges(self) -> List[Tuple[str, str]]:
+        """Get all edges as (caller, callee) tuples."""
+        edges = []
+        for callee, callers in self.callers.items():
+            for caller in callers:
+                edges.append((caller, callee))
+        return edges
+    
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary."""
+        return {
+            "callers": self.callers,
+            "callees": self.callees,
+            "edge_count": len(self.edges),
+        }
+
+
+@dataclass
+class RequestTiming:
+    """Single HTTP request timing."""
+    
+    url: str
+    method: str
+    start_time: float
+    first_byte_time: float
+    end_time: float
+    status_code: int
+    
+    @property
+    def ttfb_ms(self) -> float:
+        """Time to first byte in milliseconds."""
+        return (self.first_byte_time - self.start_time) * 1000
+    
+    @property
+    def total_ms(self) -> float:
+        """Total request time in milliseconds."""
+        return (self.end_time - self.start_time) * 1000
+    
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary."""
+        return {
+            "url": self.url,
+            "method": self.method,
+            "ttfb_ms": self.ttfb_ms,
+            "total_ms": self.total_ms,
+            "status_code": self.status_code,
+        }
+
+
+@dataclass
+class InvocationInfo:
+    """How the profile was invoked."""
+    
+    method: str  # "cli_direct", "profile_command", "tui"
+    flags: Dict[str, Any] = field(default_factory=dict)
+    praisonai_version: str = ""
+    python_version: str = ""
+    
+    def __post_init__(self):
+        """Set version info if not provided."""
+        if not self.python_version:
+            self.python_version = platform.python_version()
+        if not self.praisonai_version:
+            try:
+                from praisonai.version import __version__
+                self.praisonai_version = __version__
+            except ImportError:
+                self.praisonai_version = "unknown"
+    
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary."""
+        return {
+            "method": self.method,
+            "flags": self.flags,
+            "praisonai_version": self.praisonai_version,
+            "python_version": self.python_version,
+        }
+
+
+@dataclass
+class DecisionTrace:
+    """
+    Decision trace for deep profile visibility.
+    
+    Explains what decisions were made during execution.
+    """
+    
+    agent_config: str = ""           # Agent configuration chosen
+    model_selected: str = ""         # Model used
+    streaming_mode: bool = False     # Streaming vs non-streaming
+    profile_layer: int = 0           # Profile layer (0/1/2)
+    tools_enabled: List[str] = field(default_factory=list)
+    tools_disabled: List[str] = field(default_factory=list)
+    fallbacks_used: List[str] = field(default_factory=list)
+    
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary."""
+        return {
+            "agent_config": self.agent_config,
+            "model_selected": self.model_selected,
+            "streaming_mode": self.streaming_mode,
+            "profile_layer": self.profile_layer,
+            "tools_enabled": self.tools_enabled,
+            "tools_disabled": self.tools_disabled,
+            "fallbacks_used": self.fallbacks_used,
+        }
+    
+    def to_text(self) -> str:
+        """Format as human-readable text."""
+        lines = []
+        lines.append(f"  Agent Config:    {self.agent_config or 'default'}")
+        lines.append(f"  Model:           {self.model_selected or 'default'}")
+        lines.append(f"  Streaming:       {'enabled' if self.streaming_mode else 'disabled'}")
+        lines.append(f"  Profile Layer:   {self.profile_layer}")
+        if self.tools_enabled:
+            lines.append(f"  Tools Enabled:   {', '.join(self.tools_enabled)}")
+        if self.tools_disabled:
+            lines.append(f"  Tools Disabled:  {', '.join(self.tools_disabled)}")
+        if self.fallbacks_used:
+            lines.append(f"  Fallbacks:       {', '.join(self.fallbacks_used)}")
+        return "\n".join(lines)
+
+
+@dataclass
+class ModuleBreakdown:
+    """
+    Module/file breakdown for deep profile visibility.
+    
+    Groups functions by module category.
+    """
+    
+    cli_modules: List[str] = field(default_factory=list)
+    execution_modules: List[str] = field(default_factory=list)
+    agent_modules: List[str] = field(default_factory=list)
+    tool_modules: List[str] = field(default_factory=list)
+    network_modules: List[str] = field(default_factory=list)
+    third_party_modules: List[str] = field(default_factory=list)
+    
+    def to_dict(self) -> Dict[str, List[str]]:
+        """Convert to dictionary."""
+        return {
+            "cli": self.cli_modules,
+            "execution": self.execution_modules,
+            "agent": self.agent_modules,
+            "tools": self.tool_modules,
+            "network": self.network_modules,
+            "third_party": self.third_party_modules,
+        }
+
+
+@dataclass
+class ProfileReport:
+    """
+    Canonical profile report.
+    
+    VERSION 1.0 - Schema is frozen after release.
+    New fields MUST be optional with defaults.
+    """
+    
+    # Schema version (MANDATORY, NEVER CHANGES MEANING)
+    schema_version: str = SCHEMA_VERSION
+    
+    # Identity (MANDATORY)
+    run_id: str = ""
+    timestamp: str = ""  # ISO 8601
+    
+    # Invocation (MANDATORY)
+    invocation: InvocationInfo = field(default_factory=lambda: InvocationInfo(method="unknown"))
+    
+    # Timing (MANDATORY)
+    timing: TimingBreakdown = field(default_factory=TimingBreakdown)
+    
+    # Function stats (OPTIONAL, present if layer >= 1)
+    functions: Optional[List[FunctionStat]] = None
+    
+    # Call graph (OPTIONAL, present if layer >= 2)
+    call_graph: Optional[CallGraph] = None
+    
+    # Network (OPTIONAL, present if track_network)
+    network: Optional[List[RequestTiming]] = None
+    
+    # Response (MANDATORY but truncated)
+    response_preview: str = ""  # Max 500 chars
+    
+    # Decision trace (OPTIONAL, present if layer >= 2)
+    decision_trace: Optional[DecisionTrace] = None
+    
+    # Module breakdown (OPTIONAL, present if layer >= 2)
+    module_breakdown: Optional[ModuleBreakdown] = None
+    
+    def __post_init__(self):
+        """Set timestamp if not provided."""
+        if not self.timestamp:
+            self.timestamp = datetime.utcnow().isoformat() + "Z"
+    
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary for JSON serialization."""
+        result = {
+            "schema_version": self.schema_version,
+            "run_id": self.run_id,
+            "timestamp": self.timestamp,
+            "invocation": self.invocation.to_dict(),
+            "timing": self.timing.to_dict(),
+            "response_preview": self.response_preview,
+        }
+        
+        if self.functions is not None:
+            result["functions"] = [f.to_dict() for f in self.functions]
+        
+        if self.call_graph is not None:
+            result["call_graph"] = self.call_graph.to_dict()
+        
+        if self.network is not None:
+            result["network"] = [r.to_dict() for r in self.network]
+        
+        if self.decision_trace is not None:
+            result["decision_trace"] = self.decision_trace.to_dict()
+        
+        if self.module_breakdown is not None:
+            result["module_breakdown"] = self.module_breakdown.to_dict()
+        
+        return result
+    
+    def to_json(self) -> str:
+        """Convert to JSON string."""
+        import json
+        return json.dumps(self.to_dict(), indent=2, default=str)
+    
+    def to_text(self) -> str:
+        """Format as human-readable text."""
+        lines = []
+        lines.append("=" * 70)
+        lines.append("PraisonAI Profile Report")
+        lines.append("=" * 70)
+        lines.append("")
+        
+        # Identity
+        lines.append(f"Run ID:     {self.run_id}")
+        lines.append(f"Timestamp:  {self.timestamp}")
+        lines.append(f"Method:     {self.invocation.method}")
+        lines.append(f"Version:    {self.invocation.praisonai_version}")
+        lines.append("")
+        
+        # TIMELINE SECTION (new - shows ENTER → First Response)
+        lines.append("## Execution Timeline")
+        lines.append("-" * 45)
+        timeline = self.timing.to_timeline()
+        if timeline:
+            for phase_name, start_ms, duration_ms in timeline:
+                lines.append(f"  {phase_name:<25} : {duration_ms:>8.2f} ms")
+        else:
+            # Fallback to basic phases
+            if self.timing.imports_ms > 0:
+                lines.append(f"  {'Imports':<25} : {self.timing.imports_ms:>8.2f} ms")
+            if self.timing.agent_init_ms > 0:
+                lines.append(f"  {'Agent Init':<25} : {self.timing.agent_init_ms:>8.2f} ms")
+            if self.timing.execution_ms > 0:
+                lines.append(f"  {'Execution':<25} : {self.timing.execution_ms:>8.2f} ms")
+        lines.append("  " + "─" * 43)
+        
+        # TIME TO FIRST RESPONSE (TTFR) - the key metric
+        ttfr = self.timing.time_to_first_response_ms
+        lines.append(f"  {'⏱ Time to First Response':<25} : {ttfr:>8.2f} ms")
+        lines.append(f"  {'TOTAL':<25} : {self.timing.total_ms:>8.2f} ms")
+        lines.append("")
+        
+        # Timing breakdown (legacy format for compatibility)
+        lines.append("## Timing Breakdown")
+        lines.append("-" * 40)
+        if self.timing.cli_entry_ms > 0:
+            lines.append(f"  CLI Entry:      {self.timing.cli_entry_ms:>10.2f} ms")
+        if self.timing.cli_parse_ms > 0:
+            lines.append(f"  CLI Parse:      {self.timing.cli_parse_ms:>10.2f} ms")
+        if self.timing.routing_ms > 0:
+            lines.append(f"  Routing:        {self.timing.routing_ms:>10.2f} ms")
+        lines.append(f"  Imports:        {self.timing.imports_ms:>10.2f} ms")
+        lines.append(f"  Agent Init:     {self.timing.agent_init_ms:>10.2f} ms")
+        lines.append(f"  Execution:      {self.timing.execution_ms:>10.2f} ms")
+        if self.timing.network_start_ms > 0:
+            lines.append(f"  Network Start:  {self.timing.network_start_ms:>10.2f} ms")
+        if self.timing.first_token_ms > 0:
+            lines.append(f"  First Token:    {self.timing.first_token_ms:>10.2f} ms")
+        if self.timing.first_output_ms > 0:
+            lines.append(f"  First Output:   {self.timing.first_output_ms:>10.2f} ms")
+        lines.append("  ─────────────────────────────────────")
+        lines.append(f"  TOTAL:          {self.timing.total_ms:>10.2f} ms")
+        lines.append("")
+        
+        # Decision trace (deep profile only)
+        if self.decision_trace:
+            lines.append("## Decision Trace")
+            lines.append("-" * 40)
+            lines.append(self.decision_trace.to_text())
+            lines.append("")
+        
+        # Function stats
+        if self.functions:
+            lines.append("## Top Functions by Cumulative Time")
+            lines.append("-" * 70)
+            lines.append(f"{'Function':<40} {'Calls':>8} {'Cumul (ms)':>12}")
+            lines.append("-" * 70)
+            for func in self.functions[:20]:
+                name = func.name[:38] if len(func.name) > 38 else func.name
+                lines.append(f"{name:<40} {func.calls:>8} {func.cumulative_time_ms:>12.2f}")
+            lines.append("")
+        
+        # Module breakdown (deep profile only)
+        if self.module_breakdown:
+            lines.append("## Module Breakdown")
+            lines.append("-" * 40)
+            mb = self.module_breakdown
+            if mb.cli_modules:
+                lines.append(f"  CLI:        {len(mb.cli_modules)} modules")
+            if mb.execution_modules:
+                lines.append(f"  Execution:  {len(mb.execution_modules)} modules")
+            if mb.agent_modules:
+                lines.append(f"  Agent:      {len(mb.agent_modules)} modules")
+            if mb.tool_modules:
+                lines.append(f"  Tools:      {len(mb.tool_modules)} modules")
+            if mb.network_modules:
+                lines.append(f"  Network:    {len(mb.network_modules)} modules")
+            if mb.third_party_modules:
+                lines.append(f"  Third-party:{len(mb.third_party_modules)} modules")
+            lines.append("")
+        
+        # Call graph summary
+        if self.call_graph:
+            lines.append(f"## Call Graph: {len(self.call_graph.edges)} edges")
+            lines.append("")
+        
+        # Network timing
+        if self.network:
+            lines.append("## Network Requests")
+            lines.append("-" * 70)
+            for req in self.network[:10]:
+                lines.append(f"  {req.method} {req.url[:50]} - {req.total_ms:.2f}ms (TTFB: {req.ttfb_ms:.2f}ms)")
+            lines.append("")
+        
+        # Response preview
+        if self.response_preview:
+            lines.append("## Response Preview")
+            lines.append("-" * 40)
+            preview = self.response_preview[:200]
+            if len(self.response_preview) > 200:
+                preview += "..."
+            lines.append(preview)
+            lines.append("")
+        
+        lines.append("=" * 70)
+        
+        return "\n".join(lines)
+
+
+class Profiler:
+    """
+    External profiler that wraps execution.
+    
+    Execution code has ZERO awareness of profiling.
+    Profiler uses cProfile.runctx() or similar.
+    """
+    
+    def __init__(self, config: ProfilerConfig):
+        self.config = config
+        self._cprofile: Optional[cProfile.Profile] = None
+        self._timing: Dict[str, float] = {}
+        self._network_requests: List[RequestTiming] = []
+    
+    def profile_sync(
+        self,
+        request: ExecutionRequest,
+        invocation_method: str = "profile_command",
+    ) -> Tuple[ExecutionResult, ProfileReport]:
+        """
+        Profile synchronous execution.
+        
+        Args:
+            request: The execution request
+            invocation_method: How this was invoked (for report)
+        
+        Returns:
+            Tuple of (ExecutionResult, ProfileReport)
+        """
+        timing_dict: Dict[str, float] = {}
+        
+        # Layer 0: Wall-clock timing (always)
+        start = time.perf_counter()
+        
+        # Layer 1+: cProfile
+        if self.config.layer >= 1:
+            self._cprofile = cProfile.Profile()
+            self._cprofile.enable()
+        
+        try:
+            result = _execute_core(request, timing_dict)
+        finally:
+            if self._cprofile:
+                self._cprofile.disable()
+        
+        total_ms = (time.perf_counter() - start) * 1000
+        
+        # Build report
+        report = self._build_report(
+            request=request,
+            result=result,
+            timing_dict=timing_dict,
+            total_ms=total_ms,
+            invocation_method=invocation_method,
+        )
+        
+        return result, report
+    
+    def _build_report(
+        self,
+        request: ExecutionRequest,
+        result: ExecutionResult,
+        timing_dict: Dict[str, float],
+        total_ms: float,
+        invocation_method: str,
+    ) -> ProfileReport:
+        """Build the profile report from collected data."""
+        
+        # Calculate first_output_ms (TTFR) - time from start to when output is available
+        # This is imports + agent_init + execution (when output becomes available)
+        imports_ms = timing_dict.get("imports_ms", 0.0)
+        agent_init_ms = timing_dict.get("agent_init_ms", 0.0)
+        execution_ms = timing_dict.get("execution_ms", 0.0)
+        first_output_ms = imports_ms + agent_init_ms + execution_ms
+        
+        # Timing breakdown with full timeline
+        timing = TimingBreakdown(
+            total_ms=total_ms,
+            imports_ms=imports_ms,
+            agent_init_ms=agent_init_ms,
+            execution_ms=execution_ms,
+            cli_entry_ms=timing_dict.get("cli_entry_ms", 0.0),
+            cli_parse_ms=timing_dict.get("cli_parse_ms", 0.0),
+            routing_ms=timing_dict.get("routing_ms", 0.0),
+            network_start_ms=timing_dict.get("network_start_ms", 0.0),
+            first_token_ms=timing_dict.get("first_token_ms", 0.0),
+            first_output_ms=first_output_ms,
+        )
+        
+        # Invocation info
+        invocation = InvocationInfo(
+            method=invocation_method,
+            flags={
+                "layer": self.config.layer,
+                "stream": request.stream,
+                "model": request.model,
+            },
+        )
+        
+        # Function stats (layer 1+)
+        functions = None
+        if self.config.layer >= 1 and self._cprofile:
+            functions = self._extract_function_stats()
+        
+        # Call graph (layer 2)
+        call_graph = None
+        if self.config.layer >= 2 and self._cprofile:
+            call_graph = self._extract_call_graph()
+        
+        # Decision trace (layer 2 - deep profile)
+        decision_trace = None
+        if self.config.layer >= 2:
+            decision_trace = DecisionTrace(
+                agent_config=request.agent_name or "default",
+                model_selected=request.model or "default",
+                streaming_mode=request.stream,
+                profile_layer=self.config.layer,
+                tools_enabled=list(request.tools) if request.tools else [],
+                tools_disabled=[],
+                fallbacks_used=[],
+            )
+        
+        # Module breakdown (layer 2 - deep profile)
+        module_breakdown = None
+        if self.config.layer >= 2 and functions:
+            module_breakdown = self._extract_module_breakdown(functions)
+        
+        # Response preview (truncated)
+        response_preview = result.output[:500] if result.output else ""
+        
+        return ProfileReport(
+            run_id=result.run_id,
+            invocation=invocation,
+            timing=timing,
+            functions=functions,
+            call_graph=call_graph,
+            network=self._network_requests if self._network_requests else None,
+            response_preview=response_preview,
+            decision_trace=decision_trace,
+            module_breakdown=module_breakdown,
+        )
+    
+    def _extract_module_breakdown(self, functions: List[FunctionStat]) -> ModuleBreakdown:
+        """Extract module breakdown from function stats."""
+        cli_modules = set()
+        execution_modules = set()
+        agent_modules = set()
+        tool_modules = set()
+        network_modules = set()
+        third_party_modules = set()
+        
+        for func in functions:
+            file_path = func.file.lower()
+            
+            if "praisonai/cli" in file_path:
+                cli_modules.add(func.file)
+            elif "praisonai" in file_path and "execution" in file_path:
+                execution_modules.add(func.file)
+            elif "praisonaiagents" in file_path and "agent" in file_path:
+                agent_modules.add(func.file)
+            elif "praisonaiagents" in file_path and "tool" in file_path:
+                tool_modules.add(func.file)
+            elif any(x in file_path for x in ["httpx", "httpcore", "urllib", "requests", "aiohttp"]):
+                network_modules.add(func.file)
+            elif "site-packages" in file_path or not file_path.startswith("/"):
+                third_party_modules.add(func.file)
+        
+        return ModuleBreakdown(
+            cli_modules=list(cli_modules)[:20],
+            execution_modules=list(execution_modules)[:20],
+            agent_modules=list(agent_modules)[:20],
+            tool_modules=list(tool_modules)[:20],
+            network_modules=list(network_modules)[:20],
+            third_party_modules=list(third_party_modules)[:20],
+        )
+    
+    def _extract_function_stats(self) -> List[FunctionStat]:
+        """Extract function statistics from cProfile."""
+        if not self._cprofile:
+            return []
+        
+        stats = pstats.Stats(self._cprofile)
+        
+        # Sort by cumulative time
+        stats.sort_stats(self.config.sort_by)
+        
+        # Extract stats
+        function_stats = []
+        for (filename, line, name), (cc, nc, tt, ct, callers) in stats.stats.items():
+            if len(function_stats) >= MAX_FUNCTION_STATS:
+                break
+            
+            function_stats.append(FunctionStat(
+                name=name,
+                file=filename,
+                line=line,
+                calls=nc,
+                total_time_ms=tt * 1000,
+                cumulative_time_ms=ct * 1000,
+            ))
+        
+        # Sort and limit
+        function_stats.sort(key=lambda x: x.cumulative_time_ms, reverse=True)
+        return function_stats[:self.config.limit]
+    
+    def _extract_call_graph(self) -> CallGraph:
+        """Extract call graph from cProfile."""
+        if not self._cprofile:
+            return CallGraph()
+        
+        stats = pstats.Stats(self._cprofile)
+        
+        callers: Dict[str, List[str]] = {}
+        callees: Dict[str, List[str]] = {}
+        edge_count = 0
+        
+        for (filename, line, name), (cc, nc, tt, ct, caller_dict) in stats.stats.items():
+            if edge_count >= MAX_CALL_GRAPH_EDGES:
+                break
+            
+            callee_key = f"{name}:{filename}:{line}"
+            
+            for (caller_file, caller_line, caller_name), _ in caller_dict.items():
+                if edge_count >= MAX_CALL_GRAPH_EDGES:
+                    break
+                
+                caller_key = f"{caller_name}:{caller_file}:{caller_line}"
+                
+                # Add to callers
+                if callee_key not in callers:
+                    callers[callee_key] = []
+                if caller_key not in callers[callee_key]:
+                    callers[callee_key].append(caller_key)
+                
+                # Add to callees
+                if caller_key not in callees:
+                    callees[caller_key] = []
+                if callee_key not in callees[caller_key]:
+                    callees[caller_key].append(callee_key)
+                
+                edge_count += 1
+        
+        return CallGraph(callers=callers, callees=callees)
@@ -0,0 +1,85 @@
+"""
+ExecutionRequest - Immutable request for execution.
+
+This is the minimal, immutable request that all execution paths use.
+CLI/TUI translate their flags into this structure.
+"""
+
+from dataclasses import dataclass
+from typing import Callable, Optional, Tuple, Any
+
+
+@dataclass(frozen=True)
+class ExecutionRequest:
+    """
+    Minimal, immutable request for execution.
+    
+    NOT a kitchen-sink of all CLI flags.
+    CLI/TUI translate their flags into this.
+    
+    Attributes:
+        prompt: The user prompt to execute
+        agent_name: Name of the agent (default: "Assistant")
+        agent_instructions: Custom instructions for the agent
+        model: LLM model to use (None = use default)
+        tools: Tuple of tool callables (immutable)
+        stream: Whether to stream the response
+    
+    Note: This is frozen (immutable) to:
+    - Prevent accidental mutation during profiling
+    - Enable safe sharing across threads/coroutines
+    """
+    
+    # Required
+    prompt: str
+    
+    # Agent identity (optional, has defaults)
+    agent_name: str = "Assistant"
+    agent_instructions: Optional[str] = None
+    
+    # Model (optional, has defaults)
+    model: Optional[str] = None  # None = use default
+    
+    # Tools (optional) - using tuple for immutability
+    tools: Tuple[Callable[..., Any], ...] = ()
+    
+    # Execution mode
+    stream: bool = False
+    
+    def __post_init__(self):
+        """Validate request after initialization."""
+        if not self.prompt:
+            raise ValueError("prompt cannot be empty")
+    
+    def with_model(self, model: str) -> 'ExecutionRequest':
+        """Create a new request with a different model."""
+        return ExecutionRequest(
+            prompt=self.prompt,
+            agent_name=self.agent_name,
+            agent_instructions=self.agent_instructions,
+            model=model,
+            tools=self.tools,
+            stream=self.stream,
+        )
+    
+    def with_tools(self, tools: Tuple[Callable[..., Any], ...]) -> 'ExecutionRequest':
+        """Create a new request with different tools."""
+        return ExecutionRequest(
+            prompt=self.prompt,
+            agent_name=self.agent_name,
+            agent_instructions=self.agent_instructions,
+            model=self.model,
+            tools=tools,
+            stream=self.stream,
+        )
+    
+    def to_dict(self) -> dict:
+        """Convert to dictionary for serialization."""
+        return {
+            "prompt": self.prompt,
+            "agent_name": self.agent_name,
+            "agent_instructions": self.agent_instructions,
+            "model": self.model,
+            "tools_count": len(self.tools),
+            "stream": self.stream,
+        }
@@ -0,0 +1,74 @@
+"""
+ExecutionResult - Result from execution.
+
+Contains the output and metadata from an execution.
+"""
+
+from dataclasses import dataclass, field
+from typing import Any, Dict, List, Optional
+
+
+@dataclass
+class ExecutionResult:
+    """
+    Result from executing a request.
+    
+    Attributes:
+        output: The text output from the agent
+        run_id: Unique identifier for this execution
+        success: Whether execution completed successfully
+        error: Error message if execution failed
+        metadata: Additional metadata about the execution
+        tool_calls: List of tools that were called
+        start_time: When execution started (perf_counter)
+        end_time: When execution ended (perf_counter)
+    """
+    
+    # Primary output
+    output: str
+    
+    # Identity
+    run_id: str = ""
+    
+    # Status
+    success: bool = True
+    error: Optional[str] = None
+    
+    # Metadata
+    metadata: Dict[str, Any] = field(default_factory=dict)
+    
+    # Tool tracking
+    tool_calls: List[str] = field(default_factory=list)
+    
+    # Timing (raw perf_counter values)
+    start_time: float = 0.0
+    end_time: float = 0.0
+    
+    @property
+    def duration_ms(self) -> float:
+        """Duration in milliseconds."""
+        if self.start_time and self.end_time:
+            return (self.end_time - self.start_time) * 1000
+        return 0.0
+    
+    def to_dict(self) -> Dict[str, Any]:
+        """Convert to dictionary for serialization."""
+        return {
+            "output": self.output,
+            "run_id": self.run_id,
+            "success": self.success,
+            "error": self.error,
+            "metadata": self.metadata,
+            "tool_calls": self.tool_calls,
+            "duration_ms": self.duration_ms,
+        }
+    
+    @classmethod
+    def from_error(cls, error: str, run_id: str = "") -> 'ExecutionResult':
+        """Create a result from an error."""
+        return cls(
+            output="",
+            run_id=run_id,
+            success=False,
+            error=error,
+        )
@@ -102,6 +102,11 @@ __all__ = [
    'RuntimeConfig',
    'CodeIntelligenceRouter',
    'ActionOrchestrator',
+    # Interactive tools provider (canonical source)
+    'get_interactive_tools',
+    'resolve_tool_groups',
+    'ToolConfig',
+    'TOOL_GROUPS',
 ]

 def __getattr__(name):
@@ -247,4 +252,17 @@ def __getattr__(name):
    elif name == 'ActionOrchestrator':
        from .action_orchestrator import ActionOrchestrator
        return ActionOrchestrator
+    # Interactive tools provider (canonical source)
+    elif name == 'get_interactive_tools':
+        from .interactive_tools import get_interactive_tools
+        return get_interactive_tools
+    elif name == 'resolve_tool_groups':
+        from .interactive_tools import resolve_tool_groups
+        return resolve_tool_groups
+    elif name == 'ToolConfig':
+        from .interactive_tools import ToolConfig
+        return ToolConfig
+    elif name == 'TOOL_GROUPS':
+        from .interactive_tools import TOOL_GROUPS
+        return TOOL_GROUPS
    raise AttributeError(f"module {__name__!r} has no attribute {name!r}")
@@ -0,0 +1,380 @@
+"""
+Interactive Default Tools Provider for PraisonAI.
+
+This module provides the canonical source of truth for default tools
+in interactive modes (TUI and `praison "prompt"`).
+
+Tool Groups:
+- `acp`: ACP-powered file operations (create, edit, delete, execute)
+- `lsp`: LSP-powered code intelligence (symbols, definitions, references)
+- `basic`: Basic file tools (read, list, search)
+- `interactive`: Union of all groups (default)
+
+Usage:
+    from praisonai.cli.features.interactive_tools import (
+        get_interactive_tools,
+        resolve_tool_groups,
+        TOOL_GROUPS,
+    )
+    
+    # Get all interactive tools (default)
+    tools = get_interactive_tools()
+    
+    # Get specific groups
+    tools = get_interactive_tools(groups=["acp", "basic"])
+    
+    # Disable specific groups
+    tools = get_interactive_tools(disable=["lsp"])
+"""
+
+import logging
+import os
+from dataclasses import dataclass, field
+from typing import Callable, Dict, List, Optional, Set
+
+logger = logging.getLogger(__name__)
+
+# Tool group definitions
+TOOL_GROUPS = {
+    "acp": [
+        "acp_create_file",
+        "acp_edit_file", 
+        "acp_delete_file",
+        "acp_execute_command",
+    ],
+    "lsp": [
+        "lsp_list_symbols",
+        "lsp_find_definition",
+        "lsp_find_references",
+        "lsp_get_diagnostics",
+    ],
+    "basic": [
+        "read_file",
+        "write_file",
+        "list_files",
+        "execute_command",
+        "internet_search",
+    ],
+}
+
+# Default interactive group includes all
+TOOL_GROUPS["interactive"] = (
+    TOOL_GROUPS["acp"] + 
+    TOOL_GROUPS["lsp"] + 
+    TOOL_GROUPS["basic"]
+)
+
+
+@dataclass
+class ToolConfig:
+    """Configuration for tool loading."""
+    workspace: str = field(default_factory=os.getcwd)
+    enable_acp: bool = True
+    enable_lsp: bool = True
+    enable_basic: bool = True
+    approval_mode: str = "auto"  # auto, manual, scoped
+    lsp_enabled: bool = True
+    acp_enabled: bool = True
+    
+    @classmethod
+    def from_env(cls) -> "ToolConfig":
+        """Create config from environment variables."""
+        config = cls()
+        
+        # Check env vars for disabling groups
+        disable_str = os.environ.get("PRAISON_TOOLS_DISABLE", "")
+        if disable_str:
+            disabled = [g.strip().lower() for g in disable_str.split(",")]
+            if "acp" in disabled:
+                config.enable_acp = False
+            if "lsp" in disabled:
+                config.enable_lsp = False
+            if "basic" in disabled:
+                config.enable_basic = False
+        
+        # Check for workspace
+        workspace = os.environ.get("PRAISON_WORKSPACE", "")
+        if workspace:
+            config.workspace = workspace
+        
+        # Check approval mode
+        approval = os.environ.get("PRAISON_APPROVAL_MODE", "")
+        if approval in ("auto", "manual", "scoped"):
+            config.approval_mode = approval
+        
+        return config
+
+
+def resolve_tool_groups(
+    groups: Optional[List[str]] = None,
+    disable: Optional[List[str]] = None,
+    config: Optional[ToolConfig] = None,
+) -> Set[str]:
+    """
+    Resolve which tool names should be included based on groups and config.
+    
+    Args:
+        groups: Specific groups to include (None = all enabled)
+        disable: Groups to explicitly disable
+        config: Tool configuration
+        
+    Returns:
+        Set of tool names to include
+    """
+    if config is None:
+        config = ToolConfig.from_env()
+    
+    disable = disable or []
+    
+    # Start with requested groups or all enabled
+    if groups:
+        tool_names = set()
+        for group in groups:
+            if group in TOOL_GROUPS:
+                tool_names.update(TOOL_GROUPS[group])
+    else:
+        tool_names = set()
+        if config.enable_acp:
+            tool_names.update(TOOL_GROUPS["acp"])
+        if config.enable_lsp:
+            tool_names.update(TOOL_GROUPS["lsp"])
+        if config.enable_basic:
+            tool_names.update(TOOL_GROUPS["basic"])
+    
+    # Remove disabled groups
+    for group in disable:
+        if group in TOOL_GROUPS:
+            tool_names -= set(TOOL_GROUPS[group])
+    
+    return tool_names
+
+
+def _load_basic_tools() -> Dict[str, Callable]:
+    """Lazy load basic tools from praisonaiagents."""
+    tools = {}
+    
+    try:
+        from praisonaiagents.tools import read_file
+        tools["read_file"] = read_file
+    except ImportError:
+        logger.debug("read_file not available")
+    
+    try:
+        from praisonaiagents.tools import write_file
+        tools["write_file"] = write_file
+    except ImportError:
+        logger.debug("write_file not available")
+    
+    try:
+        from praisonaiagents.tools import list_files
+        tools["list_files"] = list_files
+    except ImportError:
+        logger.debug("list_files not available")
+    
+    try:
+        from praisonaiagents.tools import execute_command
+        tools["execute_command"] = execute_command
+    except ImportError:
+        logger.debug("execute_command not available")
+    
+    try:
+        from praisonaiagents.tools import internet_search
+        tools["internet_search"] = internet_search
+    except ImportError:
+        logger.debug("internet_search not available")
+    
+    return tools
+
+
+def _load_acp_tools(config: ToolConfig) -> Dict[str, Callable]:
+    """Lazy load ACP tools with runtime initialization."""
+    tools = {}
+    
+    try:
+        from .interactive_runtime import InteractiveRuntime, RuntimeConfig
+        from .agent_tools import create_agent_centric_tools
+        
+        # Create runtime config
+        runtime_config = RuntimeConfig(
+            workspace=config.workspace,
+            lsp_enabled=config.lsp_enabled,
+            acp_enabled=config.acp_enabled,
+            approval_mode=config.approval_mode,
+        )
+        
+        # Create runtime (lazy - doesn't start until needed)
+        runtime = InteractiveRuntime(runtime_config)
+        
+        # Get all agent-centric tools
+        all_tools = create_agent_centric_tools(runtime)
+        
+        # Map by name
+        for tool in all_tools:
+            name = tool.__name__
+            if name.startswith("acp_"):
+                tools[name] = tool
+        
+        logger.debug(f"Loaded {len(tools)} ACP tools")
+        
+    except ImportError as e:
+        logger.debug(f"ACP tools not available: {e}")
+    except Exception as e:
+        logger.warning(f"Error loading ACP tools: {e}")
+    
+    return tools
+
+
+def _load_lsp_tools(config: ToolConfig) -> Dict[str, Callable]:
+    """Lazy load LSP tools with runtime initialization."""
+    tools = {}
+    
+    try:
+        from .interactive_runtime import InteractiveRuntime, RuntimeConfig
+        from .agent_tools import create_agent_centric_tools
+        
+        # Create runtime config
+        runtime_config = RuntimeConfig(
+            workspace=config.workspace,
+            lsp_enabled=config.lsp_enabled,
+            acp_enabled=config.acp_enabled,
+            approval_mode=config.approval_mode,
+        )
+        
+        # Create runtime
+        runtime = InteractiveRuntime(runtime_config)
+        
+        # Get all agent-centric tools
+        all_tools = create_agent_centric_tools(runtime)
+        
+        # Map by name
+        for tool in all_tools:
+            name = tool.__name__
+            if name.startswith("lsp_"):
+                tools[name] = tool
+        
+        logger.debug(f"Loaded {len(tools)} LSP tools")
+        
+    except ImportError as e:
+        logger.debug(f"LSP tools not available: {e}")
+    except Exception as e:
+        logger.warning(f"Error loading LSP tools: {e}")
+    
+    return tools
+
+
+def get_interactive_tools(
+    groups: Optional[List[str]] = None,
+    disable: Optional[List[str]] = None,
+    config: Optional[ToolConfig] = None,
+    workspace: Optional[str] = None,
+) -> List[Callable]:
+    """
+    Get the default interactive tools for TUI and prompt modes.
+    
+    This is the canonical source of truth for interactive tool defaults.
+    Both `praisonai tui launch` and `praison "prompt"` should use this.
+    
+    Args:
+        groups: Specific tool groups to include (default: all enabled)
+        disable: Tool groups to disable (e.g., ["acp", "lsp"])
+        config: Tool configuration (created from env if not provided)
+        workspace: Override workspace path
+        
+    Returns:
+        List of tool callables
+        
+    Example:
+        # Default (all tools)
+        tools = get_interactive_tools()
+        
+        # Disable ACP tools
+        tools = get_interactive_tools(disable=["acp"])
+        
+        # Only basic tools
+        tools = get_interactive_tools(groups=["basic"])
+    """
+    if config is None:
+        config = ToolConfig.from_env()
+    
+    if workspace:
+        config.workspace = workspace
+    
+    # Resolve which tools to include
+    tool_names = resolve_tool_groups(groups, disable, config)
+    
+    # Load tools by group
+    all_tools: Dict[str, Callable] = {}
+    
+    # Load basic tools (always available, no runtime needed)
+    if config.enable_basic and not (disable and "basic" in disable):
+        basic_tools = _load_basic_tools()
+        all_tools.update(basic_tools)
+    
+    # Load ACP tools (requires runtime)
+    if config.enable_acp and not (disable and "acp" in disable):
+        acp_tools = _load_acp_tools(config)
+        all_tools.update(acp_tools)
+    
+    # Load LSP tools (requires runtime)
+    if config.enable_lsp and not (disable and "lsp" in disable):
+        lsp_tools = _load_lsp_tools(config)
+        all_tools.update(lsp_tools)
+    
+    # Filter to requested tools
+    result = []
+    for name in sorted(tool_names):  # Deterministic ordering
+        if name in all_tools:
+            result.append(all_tools[name])
+    
+    logger.debug(f"Loaded {len(result)} interactive tools: {[t.__name__ for t in result]}")
+    
+    return result
+
+
+def merge_tool_overrides(
+    default_tools: List[Callable],
+    add_tools: Optional[List[Callable]] = None,
+    remove_names: Optional[List[str]] = None,
+) -> List[Callable]:
+    """
+    Merge tool overrides with default tools.
+    
+    Args:
+        default_tools: Base list of tools
+        add_tools: Tools to add
+        remove_names: Tool names to remove
+        
+    Returns:
+        Merged tool list
+    """
+    # Build name -> tool map
+    tool_map = {t.__name__: t for t in default_tools}
+    
+    # Remove specified tools
+    if remove_names:
+        for name in remove_names:
+            tool_map.pop(name, None)
+    
+    # Add new tools
+    if add_tools:
+        for tool in add_tools:
+            tool_map[tool.__name__] = tool
+    
+    return list(tool_map.values())
+
+
+def get_tool_group_names() -> Dict[str, List[str]]:
+    """Get available tool groups and their tool names."""
+    return TOOL_GROUPS.copy()
+
+
+def print_tool_summary(tools: List[Callable]) -> None:
+    """Print a summary of loaded tools (for debugging)."""
+    acp_count = sum(1 for t in tools if t.__name__.startswith("acp_"))
+    lsp_count = sum(1 for t in tools if t.__name__.startswith("lsp_"))
+    basic_count = len(tools) - acp_count - lsp_count
+    
+    print(f"Interactive tools loaded: {len(tools)} total")
+    print(f"  - ACP tools: {acp_count}")
+    print(f"  - LSP tools: {lsp_count}")
+    print(f"  - Basic tools: {basic_count}")
@@ -0,0 +1,81 @@
+"""
+PraisonAI CLI Profiler Module.
+
+Provides detailed cProfile-based profiling for CLI commands and agent execution.
+Supports per-file/per-function timing, call graphs, and import time analysis.
+
+Includes Tier 0/1/2 optimizations:
+- Tier 0: Lazy imports, provider caching, CLI startup optimization
+- Tier 1: Connection pooling, prewarm hooks (opt-in)
+- Tier 2: Lite mode, async init, perf snapshot
+"""
+
+from .core import (
+    ProfilerConfig,
+    ProfilerResult,
+    QueryProfiler,
+    run_profiled_query,
+    format_profile_report,
+)
+
+from .suite import (
+    ScenarioConfig,
+    ScenarioResult,
+    SuiteResult,
+    ProfileSuiteRunner,
+    run_profile_suite,
+)
+
+from .optimizations import (
+    # Tier 0
+    ProviderCache,
+    get_provider_cache,
+    resolve_model_provider,
+    LazyImporter,
+    # Tier 1
+    ClientPool,
+    get_client_pool,
+    PrewarmManager,
+    # Tier 2
+    LiteModeConfig,
+    get_lite_mode_config,
+    is_lite_mode,
+    PerfSnapshot,
+    PerfComparison,
+    PerfSnapshotManager,
+    create_snapshot_from_suite,
+    format_comparison_report,
+)
+
+__all__ = [
+    # Core profiler
+    'ProfilerConfig',
+    'ProfilerResult', 
+    'QueryProfiler',
+    'run_profiled_query',
+    'format_profile_report',
+    # Suite runner
+    'ScenarioConfig',
+    'ScenarioResult',
+    'SuiteResult',
+    'ProfileSuiteRunner',
+    'run_profile_suite',
+    # Tier 0 optimizations
+    'ProviderCache',
+    'get_provider_cache',
+    'resolve_model_provider',
+    'LazyImporter',
+    # Tier 1 optimizations
+    'ClientPool',
+    'get_client_pool',
+    'PrewarmManager',
+    # Tier 2 optimizations
+    'LiteModeConfig',
+    'get_lite_mode_config',
+    'is_lite_mode',
+    'PerfSnapshot',
+    'PerfComparison',
+    'PerfSnapshotManager',
+    'create_snapshot_from_suite',
+    'format_comparison_report',
+]
@@ -0,0 +1,558 @@
+"""
+Core profiler implementation for PraisonAI CLI.
+
+Provides cProfile-based profiling with detailed per-file/per-function timing,
+call graphs, and import time analysis.
+"""
+
+import cProfile
+import io
+import json
+import os
+import pstats
+import re
+import subprocess
+import sys
+import time
+from dataclasses import dataclass, field
+from datetime import datetime
+from pathlib import Path
+from typing import Any, Dict, List, Optional, Tuple
+
+
+# Patterns for secrets to redact
+SECRET_PATTERNS = [
+    r'sk-[a-zA-Z0-9]{20,}',  # OpenAI keys
+    r'sk-ant-[a-zA-Z0-9-]{20,}',  # Anthropic keys
+    r'AIza[a-zA-Z0-9_-]{35}',  # Google API keys
+    r'tvly-[a-zA-Z0-9-]{20,}',  # Tavily keys
+    r'[a-zA-Z0-9_-]{32,}',  # Generic long tokens (be careful)
+]
+
+
+def redact_secrets(text: str) -> str:
+    """Redact potential secrets from text."""
+    result = text
+    for pattern in SECRET_PATTERNS[:4]:  # Skip generic pattern for safety
+        result = re.sub(pattern, '[REDACTED]', result)
+    return result
+
+
+@dataclass
+class ProfilerConfig:
+    """Configuration for profiler."""
+    deep: bool = False  # Enable deep call tracing
+    limit: int = 30  # Top N functions to show
+    sort_by: str = "cumulative"  # cumulative or tottime
+    show_files: bool = False  # Group by file
+    show_callers: bool = False  # Show callers
+    show_callees: bool = False  # Show callees
+    importtime: bool = False  # Show import timing
+    first_token: bool = False  # Track time to first token
+    save_path: Optional[str] = None  # Path to save artifacts
+    output_format: str = "text"  # text or json
+    stream: bool = False  # Streaming mode
+
+
+@dataclass
+class TimingBreakdown:
+    """Timing breakdown for profiling."""
+    cli_parse_ms: float = 0.0
+    imports_ms: float = 0.0
+    agent_construction_ms: float = 0.0
+    model_init_ms: float = 0.0
+    first_token_ms: float = 0.0
+    total_run_ms: float = 0.0
+
+
+@dataclass
+class FunctionStats:
+    """Statistics for a single function."""
+    name: str
+    filename: str
+    lineno: int
+    calls: int
+    tottime: float  # Time in function excluding subcalls
+    cumtime: float  # Time in function including subcalls
+    
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "name": self.name,
+            "filename": self.filename,
+            "lineno": self.lineno,
+            "calls": self.calls,
+            "tottime_ms": self.tottime * 1000,
+            "cumtime_ms": self.cumtime * 1000,
+        }
+
+
+@dataclass
+class FileStats:
+    """Statistics aggregated by file."""
+    filename: str
+    total_time: float
+    functions: List[FunctionStats] = field(default_factory=list)
+    
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "filename": self.filename,
+            "total_time_ms": self.total_time * 1000,
+            "function_count": len(self.functions),
+        }
+
+
+@dataclass
+class ProfilerResult:
+    """Result from a profiling run."""
+    prompt: str
+    response: str
+    timing: TimingBreakdown
+    function_stats: List[FunctionStats]
+    file_stats: List[FileStats]
+    callers: Dict[str, List[str]] = field(default_factory=dict)
+    callees: Dict[str, List[str]] = field(default_factory=dict)
+    import_times: List[Tuple[str, float]] = field(default_factory=list)
+    metadata: Dict[str, Any] = field(default_factory=dict)
+    timestamp: str = field(default_factory=lambda: datetime.utcnow().isoformat() + "Z")
+    
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "timestamp": self.timestamp,
+            "metadata": self.metadata,
+            "prompt": self.prompt[:200] + "..." if len(self.prompt) > 200 else self.prompt,
+            "response_preview": self.response[:200] + "..." if len(self.response) > 200 else self.response,
+            "timing": {
+                "cli_parse_ms": self.timing.cli_parse_ms,
+                "imports_ms": self.timing.imports_ms,
+                "agent_construction_ms": self.timing.agent_construction_ms,
+                "model_init_ms": self.timing.model_init_ms,
+                "first_token_ms": self.timing.first_token_ms,
+                "total_run_ms": self.timing.total_run_ms,
+            },
+            "top_functions": [f.to_dict() for f in self.function_stats[:20]],
+            "top_files": [f.to_dict() for f in self.file_stats[:10]],
+        }
+
+
+class QueryProfiler:
+    """
+    Profiler for query execution.
+    
+    Uses cProfile for function-level profiling and optional sys.setprofile
+    for deep call tracing.
+    """
+    
+    def __init__(self, config: Optional[ProfilerConfig] = None):
+        self.config = config or ProfilerConfig()
+        self.profiler = cProfile.Profile()
+        self._first_token_time: Optional[float] = None
+        self._start_time: float = 0.0
+        self._deep_trace_data: Dict[str, Dict] = {}
+    
+    def _on_first_token(self):
+        """Called when first token is received (for streaming)."""
+        if self._first_token_time is None:
+            self._first_token_time = time.perf_counter()
+    
+    def _deep_trace_callback(self, frame, event, arg):
+        """Callback for sys.setprofile deep tracing."""
+        if event not in ('call', 'return'):
+            return
+        
+        filename = frame.f_code.co_filename
+        funcname = frame.f_code.co_name
+        lineno = frame.f_lineno
+        
+        key = f"{filename}:{funcname}:{lineno}"
+        
+        if key not in self._deep_trace_data:
+            self._deep_trace_data[key] = {
+                'filename': filename,
+                'funcname': funcname,
+                'lineno': lineno,
+                'calls': 0,
+                'total_time': 0.0,
+                'start_times': [],
+            }
+        
+        if event == 'call':
+            self._deep_trace_data[key]['calls'] += 1
+            self._deep_trace_data[key]['start_times'].append(time.perf_counter())
+        elif event == 'return' and self._deep_trace_data[key]['start_times']:
+            start = self._deep_trace_data[key]['start_times'].pop()
+            self._deep_trace_data[key]['total_time'] += time.perf_counter() - start
+    
+    def profile_query(
+        self,
+        prompt: str,
+        model: Optional[str] = None,
+        stream: bool = False,
+    ) -> ProfilerResult:
+        """
+        Profile a query execution.
+        
+        Args:
+            prompt: The prompt to execute
+            model: Optional model to use
+            stream: Whether to use streaming mode
+            
+        Returns:
+            ProfilerResult with detailed timing and statistics
+        """
+        timing = TimingBreakdown()
+        response = ""
+        
+        # Metadata
+        metadata = self._collect_metadata(model)
+        
+        # Time CLI parsing (simulated - actual parsing happens before this)
+        cli_start = time.perf_counter()
+        timing.cli_parse_ms = (time.perf_counter() - cli_start) * 1000
+        
+        # Time imports
+        import_start = time.perf_counter()
+        try:
+            from praisonaiagents import Agent
+        except ImportError:
+            raise ImportError("praisonaiagents not installed")
+        timing.imports_ms = (time.perf_counter() - import_start) * 1000
+        
+        # Time agent construction
+        construct_start = time.perf_counter()
+        agent_config = {
+            "name": "ProfilerAgent",
+            "role": "Assistant",
+            "goal": "Complete the task",
+            "verbose": False,
+        }
+        if model:
+            agent_config["llm"] = model
+        
+        agent = Agent(**agent_config)
+        timing.agent_construction_ms = (time.perf_counter() - construct_start) * 1000
+        
+        # Time model initialization (first call)
+        model_init_start = time.perf_counter()
+        timing.model_init_ms = (time.perf_counter() - model_init_start) * 1000
+        
+        # Profile the actual execution
+        self._start_time = time.perf_counter()
+        self._first_token_time = None
+        
+        if self.config.deep:
+            sys.setprofile(self._deep_trace_callback)
+        
+        self.profiler.enable()
+        
+        try:
+            if stream and hasattr(agent, '_start_stream'):
+                # Streaming mode with first token tracking
+                chunks = []
+                for chunk in agent._start_stream(prompt):
+                    if not chunks:
+                        self._on_first_token()
+                    chunks.append(chunk)
+                response = ''.join(chunks)
+            else:
+                # Non-streaming mode
+                response = agent.start(prompt)
+                if response is None:
+                    response = ""
+        finally:
+            self.profiler.disable()
+            if self.config.deep:
+                sys.setprofile(None)
+        
+        end_time = time.perf_counter()
+        timing.total_run_ms = (end_time - self._start_time) * 1000
+        
+        if self._first_token_time:
+            timing.first_token_ms = (self._first_token_time - self._start_time) * 1000
+        
+        # Extract statistics
+        function_stats = self._extract_function_stats()
+        file_stats = self._extract_file_stats(function_stats)
+        callers, callees = self._extract_call_graph()
+        
+        # Get import times if requested
+        import_times = []
+        if self.config.importtime:
+            import_times = self._get_import_times()
+        
+        return ProfilerResult(
+            prompt=prompt,
+            response=str(response),
+            timing=timing,
+            function_stats=function_stats,
+            file_stats=file_stats,
+            callers=callers if self.config.show_callers else {},
+            callees=callees if self.config.show_callees else {},
+            import_times=import_times,
+            metadata=metadata,
+        )
+    
+    def _collect_metadata(self, model: Optional[str]) -> Dict[str, Any]:
+        """Collect system metadata."""
+        import platform
+        try:
+            from praisonai.version import __version__ as praisonai_version
+        except ImportError:
+            praisonai_version = "unknown"
+        
+        return {
+            "python_version": platform.python_version(),
+            "platform": platform.platform(),
+            "praisonai_version": praisonai_version,
+            "model": model or "default",
+            "timestamp": datetime.utcnow().isoformat() + "Z",
+        }
+    
+    def _extract_function_stats(self) -> List[FunctionStats]:
+        """Extract function statistics from profiler."""
+        stats_stream = io.StringIO()
+        stats = pstats.Stats(self.profiler, stream=stats_stream)
+        
+        # Sort by configured key
+        sort_key = pstats.SortKey.CUMULATIVE if self.config.sort_by == "cumulative" else pstats.SortKey.TIME
+        stats.sort_stats(sort_key)
+        
+        function_stats = []
+        for (filename, lineno, funcname), (cc, nc, tt, ct, callers) in stats.stats.items():
+            function_stats.append(FunctionStats(
+                name=funcname,
+                filename=filename,
+                lineno=lineno,
+                calls=nc,
+                tottime=tt,
+                cumtime=ct,
+            ))
+        
+        # Sort and limit
+        if self.config.sort_by == "cumulative":
+            function_stats.sort(key=lambda x: x.cumtime, reverse=True)
+        else:
+            function_stats.sort(key=lambda x: x.tottime, reverse=True)
+        
+        return function_stats[:self.config.limit]
+    
+    def _extract_file_stats(self, function_stats: List[FunctionStats]) -> List[FileStats]:
+        """Aggregate statistics by file."""
+        file_map: Dict[str, FileStats] = {}
+        
+        for func in function_stats:
+            if func.filename not in file_map:
+                file_map[func.filename] = FileStats(
+                    filename=func.filename,
+                    total_time=0.0,
+                    functions=[],
+                )
+            file_map[func.filename].total_time += func.cumtime
+            file_map[func.filename].functions.append(func)
+        
+        file_stats = list(file_map.values())
+        file_stats.sort(key=lambda x: x.total_time, reverse=True)
+        
+        return file_stats[:self.config.limit]
+    
+    def _extract_call_graph(self) -> Tuple[Dict[str, List[str]], Dict[str, List[str]]]:
+        """Extract caller/callee relationships."""
+        callers: Dict[str, List[str]] = {}
+        callees: Dict[str, List[str]] = {}
+        
+        stats = pstats.Stats(self.profiler)
+        
+        for (filename, lineno, funcname), (cc, nc, tt, ct, caller_dict) in stats.stats.items():
+            func_key = f"{funcname} ({os.path.basename(filename)}:{lineno})"
+            
+            if caller_dict:
+                callers[func_key] = []
+                for (caller_file, caller_line, caller_name), _ in caller_dict.items():
+                    caller_key = f"{caller_name} ({os.path.basename(caller_file)}:{caller_line})"
+                    callers[func_key].append(caller_key)
+                    
+                    if caller_key not in callees:
+                        callees[caller_key] = []
+                    callees[caller_key].append(func_key)
+        
+        return callers, callees
+    
+    def _get_import_times(self) -> List[Tuple[str, float]]:
+        """Get import timing by running subprocess with -X importtime."""
+        try:
+            result = subprocess.run(
+                [sys.executable, "-X", "importtime", "-c", "import praisonaiagents"],
+                capture_output=True,
+                text=True,
+                timeout=30,
+            )
+            
+            import_times = []
+            for line in result.stderr.split('\n'):
+                if 'import time:' in line:
+                    # Parse: "import time:       123 |        456 | module_name"
+                    match = re.search(r'import time:\s+(\d+)\s+\|\s+(\d+)\s+\|\s+(.+)', line)
+                    if match:
+                        self_time = int(match.group(1))
+                        cumulative = int(match.group(2))
+                        module = match.group(3).strip()
+                        import_times.append((module, cumulative / 1000000))  # Convert to seconds
+            
+            # Sort by time descending
+            import_times.sort(key=lambda x: x[1], reverse=True)
+            return import_times[:20]  # Top 20
+            
+        except Exception:
+            return []
+    
+    def save_artifacts(self, result: ProfilerResult, base_path: str):
+        """Save profiling artifacts to files."""
+        base = Path(base_path)
+        base.parent.mkdir(parents=True, exist_ok=True)
+        
+        # Save .prof binary
+        prof_path = str(base) + ".prof"
+        self.profiler.dump_stats(prof_path)
+        
+        # Save .txt report
+        txt_path = str(base) + ".txt"
+        with open(txt_path, 'w') as f:
+            f.write(format_profile_report(result, self.config))
+        
+        # Save .json if requested
+        if self.config.output_format == "json":
+            json_path = str(base) + ".json"
+            with open(json_path, 'w') as f:
+                json.dump(result.to_dict(), f, indent=2, default=str)
+        
+        return prof_path, txt_path
+
+
+def run_profiled_query(
+    prompt: str,
+    config: Optional[ProfilerConfig] = None,
+    model: Optional[str] = None,
+) -> ProfilerResult:
+    """
+    Run a profiled query.
+    
+    Args:
+        prompt: The prompt to execute
+        config: Profiler configuration
+        model: Optional model to use
+        
+    Returns:
+        ProfilerResult with detailed timing and statistics
+    """
+    profiler = QueryProfiler(config)
+    return profiler.profile_query(prompt, model=model, stream=config.stream if config else False)
+
+
+def format_profile_report(result: ProfilerResult, config: Optional[ProfilerConfig] = None) -> str:
+    """
+    Format a profile result as a text report.
+    
+    Args:
+        result: ProfilerResult to format
+        config: Optional profiler config
+        
+    Returns:
+        Formatted text report
+    """
+    config = config or ProfilerConfig()
+    lines = []
+    
+    # Header
+    lines.append("=" * 70)
+    lines.append("PraisonAI Profile Report")
+    lines.append("=" * 70)
+    lines.append("")
+    
+    # Metadata
+    lines.append("## System Information")
+    lines.append(f"  Timestamp:        {result.timestamp}")
+    lines.append(f"  Python Version:   {result.metadata.get('python_version', 'N/A')}")
+    lines.append(f"  Platform:         {result.metadata.get('platform', 'N/A')}")
+    lines.append(f"  PraisonAI:        {result.metadata.get('praisonai_version', 'N/A')}")
+    lines.append(f"  Model:            {result.metadata.get('model', 'N/A')}")
+    lines.append("")
+    
+    # Timing breakdown
+    lines.append("## Timing Breakdown")
+    lines.append(f"  CLI Parse:        {result.timing.cli_parse_ms:>10.2f} ms")
+    lines.append(f"  Imports:          {result.timing.imports_ms:>10.2f} ms")
+    lines.append(f"  Agent Construct:  {result.timing.agent_construction_ms:>10.2f} ms")
+    lines.append(f"  Model Init:       {result.timing.model_init_ms:>10.2f} ms")
+    if result.timing.first_token_ms > 0:
+        lines.append(f"  First Token:      {result.timing.first_token_ms:>10.2f} ms")
+    lines.append(f"  Total Run:        {result.timing.total_run_ms:>10.2f} ms")
+    lines.append("")
+    
+    # Per-file timing (if requested)
+    if config.show_files and result.file_stats:
+        lines.append("## Per-File Timing (Top Files)")
+        lines.append("-" * 70)
+        lines.append(f"{'File':<50} {'Time (ms)':>15}")
+        lines.append("-" * 70)
+        for fs in result.file_stats[:15]:
+            filename = os.path.basename(fs.filename)
+            if len(filename) > 48:
+                filename = "..." + filename[-45:]
+            lines.append(f"{filename:<50} {fs.total_time * 1000:>15.2f}")
+        lines.append("")
+    
+    # Per-function timing
+    lines.append("## Per-Function Timing (Top Functions)")
+    lines.append("-" * 70)
+    sort_label = "Cumulative" if config.sort_by == "cumulative" else "Total"
+    lines.append(f"{'Function':<35} {'Calls':>8} {sort_label + ' (ms)':>12} {'Self (ms)':>12}")
+    lines.append("-" * 70)
+    for fs in result.function_stats[:config.limit]:
+        funcname = fs.name
+        if len(funcname) > 33:
+            funcname = funcname[:30] + "..."
+        lines.append(f"{funcname:<35} {fs.calls:>8} {fs.cumtime * 1000:>12.2f} {fs.tottime * 1000:>12.2f}")
+    lines.append("")
+    
+    # Callers (if requested)
+    if config.show_callers and result.callers:
+        lines.append("## Callers (Who called each function)")
+        lines.append("-" * 70)
+        for func, caller_list in list(result.callers.items())[:10]:
+            lines.append(f"  {func}:")
+            for caller in caller_list[:5]:
+                lines.append(f"    <- {caller}")
+        lines.append("")
+    
+    # Callees (if requested)
+    if config.show_callees and result.callees:
+        lines.append("## Callees (What each function called)")
+        lines.append("-" * 70)
+        for func, callee_list in list(result.callees.items())[:10]:
+            lines.append(f"  {func}:")
+            for callee in callee_list[:5]:
+                lines.append(f"    -> {callee}")
+        lines.append("")
+    
+    # Import times (if requested)
+    if config.importtime and result.import_times:
+        lines.append("## Import Times (Top Modules)")
+        lines.append("-" * 70)
+        lines.append(f"{'Module':<50} {'Time (ms)':>15}")
+        lines.append("-" * 70)
+        for module, time_sec in result.import_times[:15]:
+            module_name = module
+            if len(module_name) > 48:
+                module_name = "..." + module_name[-45:]
+            lines.append(f"{module_name:<50} {time_sec * 1000:>15.2f}")
+        lines.append("")
+    
+    # Response preview
+    lines.append("## Response Preview")
+    lines.append("-" * 70)
+    preview = result.response[:500] + "..." if len(result.response) > 500 else result.response
+    lines.append(preview)
+    lines.append("")
+    
+    lines.append("=" * 70)
+    
+    return "\n".join(lines)
@@ -0,0 +1,652 @@
+"""
+Performance Optimizations for PraisonAI CLI.
+
+Implements Tier 0/1/2 optimizations:
+- Tier 0: Lazy imports, provider caching, CLI startup optimization
+- Tier 1: Connection pooling, prewarm hooks (opt-in)
+- Tier 2: Lite mode, async init, perf snapshot
+
+All optimizations are safe-by-default and opt-in where behavior changes.
+Multi-agent safe: no global mutable state that affects concurrent runs.
+"""
+
+import os
+import json
+import threading
+import time
+from dataclasses import dataclass, field
+from pathlib import Path
+from typing import Any, Callable, Dict, List, Optional, Tuple
+from functools import lru_cache
+
+
+# =============================================================================
+# TIER 0: Safe Fast Wins (No behavior change)
+# =============================================================================
+
+class ProviderCache:
+    """
+    Thread-safe cache for provider/model resolution.
+    
+    Caches resolved provider configurations to avoid repeated lookups.
+    Per-process cache with clear invalidation rules.
+    Multi-agent safe: uses thread-local storage for mutable state.
+    """
+    
+    _instance = None
+    _lock = threading.Lock()
+    
+    def __new__(cls):
+        if cls._instance is None:
+            with cls._lock:
+                if cls._instance is None:
+                    cls._instance = super().__new__(cls)
+                    cls._instance._cache = {}
+                    cls._instance._cache_lock = threading.Lock()
+                    cls._instance._hits = 0
+                    cls._instance._misses = 0
+        return cls._instance
+    
+    def get(self, key: str) -> Optional[Any]:
+        """Get cached value (thread-safe)."""
+        with self._cache_lock:
+            if key in self._cache:
+                self._hits += 1
+                return self._cache[key]
+            self._misses += 1
+            return None
+    
+    def set(self, key: str, value: Any, ttl_seconds: int = 300) -> None:
+        """Set cached value with TTL (thread-safe)."""
+        with self._cache_lock:
+            self._cache[key] = {
+                'value': value,
+                'expires': time.time() + ttl_seconds,
+            }
+    
+    def invalidate(self, key: Optional[str] = None) -> None:
+        """Invalidate cache entry or entire cache."""
+        with self._cache_lock:
+            if key:
+                self._cache.pop(key, None)
+            else:
+                self._cache.clear()
+    
+    def get_stats(self) -> Dict[str, int]:
+        """Get cache statistics."""
+        with self._cache_lock:
+            return {
+                'hits': self._hits,
+                'misses': self._misses,
+                'size': len(self._cache),
+            }
+    
+    def cleanup_expired(self) -> int:
+        """Remove expired entries. Returns count of removed entries."""
+        now = time.time()
+        removed = 0
+        with self._cache_lock:
+            expired_keys = [
+                k for k, v in self._cache.items()
+                if isinstance(v, dict) and v.get('expires', float('inf')) < now
+            ]
+            for k in expired_keys:
+                del self._cache[k]
+                removed += 1
+        return removed
+
+
+# Global provider cache instance (singleton, thread-safe)
+_provider_cache = None
+
+def get_provider_cache() -> ProviderCache:
+    """Get the global provider cache instance."""
+    global _provider_cache
+    if _provider_cache is None:
+        _provider_cache = ProviderCache()
+    return _provider_cache
+
+
+@lru_cache(maxsize=32)
+def resolve_model_provider(model: str) -> Tuple[str, str]:
+    """
+    Resolve model string to (provider, model_name) tuple.
+    
+    Cached using LRU cache for repeated lookups.
+    """
+    if '/' in model:
+        parts = model.split('/', 1)
+        return parts[0], parts[1]
+    
+    # Default provider mappings
+    if model.startswith('gpt-') or model.startswith('o1-') or model.startswith('o3-'):
+        return 'openai', model
+    elif model.startswith('claude-'):
+        return 'anthropic', model
+    elif model.startswith('gemini-'):
+        return 'google', model
+    elif model.startswith('llama') or model.startswith('mistral'):
+        return 'ollama', model
+    
+    return 'openai', model  # Default to OpenAI
+
+
+class LazyImporter:
+    """
+    Lazy importer for heavy modules.
+    
+    Defers import until first use, reducing CLI startup time.
+    Thread-safe and caches imported modules.
+    """
+    
+    _imports: Dict[str, Any] = {}
+    _lock = threading.Lock()
+    
+    @classmethod
+    def get(cls, module_path: str, attr: Optional[str] = None) -> Any:
+        """
+        Lazily import a module or attribute.
+        
+        Args:
+            module_path: Full module path (e.g., 'openai.OpenAI')
+            attr: Optional attribute to get from module
+            
+        Returns:
+            Imported module or attribute
+        """
+        cache_key = f"{module_path}.{attr}" if attr else module_path
+        
+        with cls._lock:
+            if cache_key in cls._imports:
+                return cls._imports[cache_key]
+        
+        # Import outside lock to avoid deadlocks
+        import importlib
+        
+        if attr:
+            module = importlib.import_module(module_path)
+            result = getattr(module, attr)
+        else:
+            result = importlib.import_module(module_path)
+        
+        with cls._lock:
+            cls._imports[cache_key] = result
+        
+        return result
+    
+    @classmethod
+    def clear(cls) -> None:
+        """Clear import cache."""
+        with cls._lock:
+            cls._imports.clear()
+
+
+# =============================================================================
+# TIER 1: Medium Effort (Guarded/opt-in)
+# =============================================================================
+
+class ClientPool:
+    """
+    Connection pool for API clients.
+    
+    Reuses client instances for repeated calls to the same provider.
+    Multi-agent safe: clients are isolated by key.
+    Does not leak API keys between different configurations.
+    """
+    
+    def __init__(self, max_size: int = 10):
+        self._pool: Dict[str, Any] = {}
+        self._lock = threading.Lock()
+        self._max_size = max_size
+        self._access_times: Dict[str, float] = {}
+    
+    def get_or_create(
+        self,
+        key: str,
+        factory: Callable[[], Any],
+    ) -> Any:
+        """
+        Get existing client or create new one.
+        
+        Args:
+            key: Unique key for this client configuration
+            factory: Function to create new client if needed
+            
+        Returns:
+            Client instance
+        """
+        with self._lock:
+            if key in self._pool:
+                self._access_times[key] = time.time()
+                return self._pool[key]
+            
+            # Evict oldest if at capacity
+            if len(self._pool) >= self._max_size:
+                oldest_key = min(self._access_times, key=self._access_times.get)
+                del self._pool[oldest_key]
+                del self._access_times[oldest_key]
+        
+        # Create outside lock
+        client = factory()
+        
+        with self._lock:
+            self._pool[key] = client
+            self._access_times[key] = time.time()
+        
+        return client
+    
+    def remove(self, key: str) -> None:
+        """Remove client from pool."""
+        with self._lock:
+            self._pool.pop(key, None)
+            self._access_times.pop(key, None)
+    
+    def clear(self) -> None:
+        """Clear all clients from pool."""
+        with self._lock:
+            self._pool.clear()
+            self._access_times.clear()
+    
+    def size(self) -> int:
+        """Get current pool size."""
+        with self._lock:
+            return len(self._pool)
+
+
+# Global client pool (opt-in usage)
+_client_pool: Optional[ClientPool] = None
+
+def get_client_pool() -> ClientPool:
+    """Get the global client pool instance."""
+    global _client_pool
+    if _client_pool is None:
+        _client_pool = ClientPool()
+    return _client_pool
+
+
+class PrewarmManager:
+    """
+    Manager for pre-warming provider connections.
+    
+    OPT-IN ONLY: Must be explicitly enabled.
+    Pre-initializes provider clients in background to reduce first-call latency.
+    """
+    
+    _enabled = False
+    _prewarmed: Dict[str, bool] = {}
+    _lock = threading.Lock()
+    _background_thread: Optional[threading.Thread] = None
+    
+    @classmethod
+    def enable(cls) -> None:
+        """Enable pre-warming (opt-in)."""
+        cls._enabled = True
+    
+    @classmethod
+    def disable(cls) -> None:
+        """Disable pre-warming."""
+        cls._enabled = False
+    
+    @classmethod
+    def is_enabled(cls) -> bool:
+        """Check if pre-warming is enabled."""
+        return cls._enabled
+    
+    @classmethod
+    def prewarm_provider(cls, provider: str, api_key: Optional[str] = None) -> None:
+        """
+        Pre-warm a provider connection in background.
+        
+        Only runs if pre-warming is enabled.
+        """
+        if not cls._enabled:
+            return
+        
+        with cls._lock:
+            if provider in cls._prewarmed:
+                return
+            cls._prewarmed[provider] = False
+        
+        def _prewarm():
+            try:
+                if provider == 'openai':
+                    # Just import and create client - don't make API call
+                    from openai import OpenAI
+                    key = api_key or os.environ.get('OPENAI_API_KEY')
+                    if key:
+                        _ = OpenAI(api_key=key)
+                elif provider == 'anthropic':
+                    from anthropic import Anthropic
+                    key = api_key or os.environ.get('ANTHROPIC_API_KEY')
+                    if key:
+                        _ = Anthropic(api_key=key)
+                
+                with cls._lock:
+                    cls._prewarmed[provider] = True
+            except Exception:
+                pass  # Silently fail - pre-warming is optional
+        
+        # Run in background thread
+        thread = threading.Thread(target=_prewarm, daemon=True)
+        thread.start()
+    
+    @classmethod
+    def is_prewarmed(cls, provider: str) -> bool:
+        """Check if provider is pre-warmed."""
+        with cls._lock:
+            return cls._prewarmed.get(provider, False)
+    
+    @classmethod
+    def reset(cls) -> None:
+        """Reset pre-warm state."""
+        with cls._lock:
+            cls._prewarmed.clear()
+            cls._enabled = False
+
+
+# =============================================================================
+# TIER 2: Architectural (Safe and non-breaking)
+# =============================================================================
+
+@dataclass
+class LiteModeConfig:
+    """
+    Configuration for lite runtime mode.
+    
+    OPT-IN ONLY: Defaults to OFF.
+    Avoids expensive type/model loading when not required.
+    """
+    enabled: bool = False
+    skip_type_validation: bool = False
+    skip_model_validation: bool = False
+    minimal_imports: bool = False
+    
+    @classmethod
+    def from_env(cls) -> 'LiteModeConfig':
+        """Create config from environment variables."""
+        return cls(
+            enabled=os.environ.get('PRAISONAI_LITE_MODE', '').lower() in ('1', 'true', 'yes'),
+            skip_type_validation=os.environ.get('PRAISONAI_SKIP_TYPE_VALIDATION', '').lower() in ('1', 'true'),
+            skip_model_validation=os.environ.get('PRAISONAI_SKIP_MODEL_VALIDATION', '').lower() in ('1', 'true'),
+            minimal_imports=os.environ.get('PRAISONAI_MINIMAL_IMPORTS', '').lower() in ('1', 'true'),
+        )
+
+
+# Global lite mode config
+_lite_mode_config: Optional[LiteModeConfig] = None
+
+def get_lite_mode_config() -> LiteModeConfig:
+    """Get lite mode configuration."""
+    global _lite_mode_config
+    if _lite_mode_config is None:
+        _lite_mode_config = LiteModeConfig.from_env()
+    return _lite_mode_config
+
+def is_lite_mode() -> bool:
+    """Check if lite mode is enabled."""
+    return get_lite_mode_config().enabled
+
+
+@dataclass
+class PerfSnapshot:
+    """
+    Performance snapshot for baseline comparison.
+    
+    Records timing data that can be compared against later runs.
+    """
+    timestamp: str
+    name: str
+    startup_cold_ms: float
+    startup_warm_ms: float
+    import_time_ms: float
+    query_time_ms: float
+    first_token_ms: float
+    metadata: Dict[str, Any] = field(default_factory=dict)
+    
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            'timestamp': self.timestamp,
+            'name': self.name,
+            'startup_cold_ms': self.startup_cold_ms,
+            'startup_warm_ms': self.startup_warm_ms,
+            'import_time_ms': self.import_time_ms,
+            'query_time_ms': self.query_time_ms,
+            'first_token_ms': self.first_token_ms,
+            'metadata': self.metadata,
+        }
+    
+    @classmethod
+    def from_dict(cls, data: Dict[str, Any]) -> 'PerfSnapshot':
+        return cls(
+            timestamp=data['timestamp'],
+            name=data['name'],
+            startup_cold_ms=data['startup_cold_ms'],
+            startup_warm_ms=data['startup_warm_ms'],
+            import_time_ms=data['import_time_ms'],
+            query_time_ms=data['query_time_ms'],
+            first_token_ms=data.get('first_token_ms', 0.0),
+            metadata=data.get('metadata', {}),
+        )
+
+
+@dataclass
+class PerfComparison:
+    """Comparison between two performance snapshots."""
+    baseline: PerfSnapshot
+    current: PerfSnapshot
+    
+    @property
+    def startup_cold_diff_ms(self) -> float:
+        return self.current.startup_cold_ms - self.baseline.startup_cold_ms
+    
+    @property
+    def startup_cold_diff_pct(self) -> float:
+        if self.baseline.startup_cold_ms == 0:
+            return 0.0
+        return (self.startup_cold_diff_ms / self.baseline.startup_cold_ms) * 100
+    
+    @property
+    def import_time_diff_ms(self) -> float:
+        return self.current.import_time_ms - self.baseline.import_time_ms
+    
+    @property
+    def import_time_diff_pct(self) -> float:
+        if self.baseline.import_time_ms == 0:
+            return 0.0
+        return (self.import_time_diff_ms / self.baseline.import_time_ms) * 100
+    
+    @property
+    def query_time_diff_ms(self) -> float:
+        return self.current.query_time_ms - self.baseline.query_time_ms
+    
+    @property
+    def query_time_diff_pct(self) -> float:
+        if self.baseline.query_time_ms == 0:
+            return 0.0
+        return (self.query_time_diff_ms / self.baseline.query_time_ms) * 100
+    
+    def is_regression(self, threshold_pct: float = 10.0) -> bool:
+        """Check if there's a performance regression above threshold."""
+        return (
+            self.startup_cold_diff_pct > threshold_pct or
+            self.import_time_diff_pct > threshold_pct or
+            self.query_time_diff_pct > threshold_pct
+        )
+    
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            'baseline': self.baseline.to_dict(),
+            'current': self.current.to_dict(),
+            'diffs': {
+                'startup_cold_ms': self.startup_cold_diff_ms,
+                'startup_cold_pct': self.startup_cold_diff_pct,
+                'import_time_ms': self.import_time_diff_ms,
+                'import_time_pct': self.import_time_diff_pct,
+                'query_time_ms': self.query_time_diff_ms,
+                'query_time_pct': self.query_time_diff_pct,
+            },
+            'is_regression': self.is_regression(),
+        }
+
+
+class PerfSnapshotManager:
+    """
+    Manager for performance snapshots.
+    
+    Stores snapshots locally for baseline comparison.
+    OPT-IN: User must explicitly invoke snapshot commands.
+    """
+    
+    DEFAULT_DIR = Path.home() / '.praisonai' / 'perf_snapshots'
+    
+    def __init__(self, snapshot_dir: Optional[Path] = None):
+        self.snapshot_dir = snapshot_dir or self.DEFAULT_DIR
+        self.snapshot_dir.mkdir(parents=True, exist_ok=True)
+    
+    def save(self, snapshot: PerfSnapshot) -> Path:
+        """Save a snapshot to disk."""
+        filename = f"{snapshot.name}_{snapshot.timestamp.replace(':', '-').replace('.', '-')}.json"
+        path = self.snapshot_dir / filename
+        with open(path, 'w') as f:
+            json.dump(snapshot.to_dict(), f, indent=2)
+        return path
+    
+    def load(self, name: str) -> Optional[PerfSnapshot]:
+        """Load the most recent snapshot with given name."""
+        pattern = f"{name}_*.json"
+        files = sorted(self.snapshot_dir.glob(pattern), reverse=True)
+        if not files:
+            return None
+        
+        with open(files[0]) as f:
+            data = json.load(f)
+        return PerfSnapshot.from_dict(data)
+    
+    def load_baseline(self, name: str = 'baseline') -> Optional[PerfSnapshot]:
+        """Load the baseline snapshot."""
+        return self.load(name)
+    
+    def save_baseline(self, snapshot: PerfSnapshot) -> Path:
+        """Save as baseline (overwrites previous baseline)."""
+        snapshot.name = 'baseline'
+        # Remove old baselines
+        for f in self.snapshot_dir.glob('baseline_*.json'):
+            f.unlink()
+        return self.save(snapshot)
+    
+    def compare(self, current: PerfSnapshot, baseline_name: str = 'baseline') -> Optional[PerfComparison]:
+        """Compare current snapshot against baseline."""
+        baseline = self.load(baseline_name)
+        if not baseline:
+            return None
+        return PerfComparison(baseline=baseline, current=current)
+    
+    def list_snapshots(self) -> List[str]:
+        """List all saved snapshot names."""
+        files = self.snapshot_dir.glob('*.json')
+        names = set()
+        for f in files:
+            # Extract name from filename (before first underscore)
+            name = f.stem.rsplit('_', 1)[0] if '_' in f.stem else f.stem
+            names.add(name)
+        return sorted(names)
+
+
+# =============================================================================
+# Utility Functions
+# =============================================================================
+
+def create_snapshot_from_suite(suite_result: Any, name: str = 'current') -> PerfSnapshot:
+    """
+    Create a PerfSnapshot from a SuiteResult.
+    
+    Args:
+        suite_result: SuiteResult from profile suite
+        name: Name for the snapshot
+        
+    Returns:
+        PerfSnapshot instance
+    """
+    # Get average query time from scenarios
+    query_times = []
+    first_token_times = []
+    for scenario in suite_result.scenarios:
+        query_times.extend(scenario.total_times)
+        first_token_times.extend(scenario.first_token_times)
+    
+    avg_query_time = sum(query_times) / len(query_times) if query_times else 0.0
+    avg_first_token = sum(first_token_times) / len(first_token_times) if first_token_times else 0.0
+    
+    # Get import time from analysis
+    import_time = 0.0
+    if suite_result.import_analysis:
+        import_time = suite_result.import_analysis[0].get('cumulative_ms', 0.0)
+    
+    return PerfSnapshot(
+        timestamp=suite_result.timestamp,
+        name=name,
+        startup_cold_ms=suite_result.startup_cold_ms,
+        startup_warm_ms=suite_result.startup_warm_ms,
+        import_time_ms=import_time,
+        query_time_ms=avg_query_time,
+        first_token_ms=avg_first_token,
+        metadata=suite_result.metadata,
+    )
+
+
+def format_comparison_report(comparison: PerfComparison) -> str:
+    """Format a comparison report as text."""
+    lines = []
+    lines.append("=" * 70)
+    lines.append("Performance Comparison Report")
+    lines.append("=" * 70)
+    lines.append("")
+    
+    lines.append(f"Baseline: {comparison.baseline.name} ({comparison.baseline.timestamp})")
+    lines.append(f"Current:  {comparison.current.name} ({comparison.current.timestamp})")
+    lines.append("")
+    
+    lines.append("-" * 70)
+    lines.append(f"{'Metric':<25} {'Baseline':>12} {'Current':>12} {'Diff':>12} {'%':>8}")
+    lines.append("-" * 70)
+    
+    def format_row(name: str, baseline: float, current: float, diff: float, pct: float) -> str:
+        sign = '+' if diff > 0 else ''
+        pct_sign = '+' if pct > 0 else ''
+        return f"{name:<25} {baseline:>12.2f} {current:>12.2f} {sign}{diff:>11.2f} {pct_sign}{pct:>7.1f}%"
+    
+    lines.append(format_row(
+        "Startup Cold (ms)",
+        comparison.baseline.startup_cold_ms,
+        comparison.current.startup_cold_ms,
+        comparison.startup_cold_diff_ms,
+        comparison.startup_cold_diff_pct,
+    ))
+    
+    lines.append(format_row(
+        "Import Time (ms)",
+        comparison.baseline.import_time_ms,
+        comparison.current.import_time_ms,
+        comparison.import_time_diff_ms,
+        comparison.import_time_diff_pct,
+    ))
+    
+    lines.append(format_row(
+        "Query Time (ms)",
+        comparison.baseline.query_time_ms,
+        comparison.current.query_time_ms,
+        comparison.query_time_diff_ms,
+        comparison.query_time_diff_pct,
+    ))
+    
+    lines.append("-" * 70)
+    lines.append("")
+    
+    if comparison.is_regression():
+        lines.append("⚠️  REGRESSION DETECTED (>10% slower)")
+    else:
+        lines.append("✅ No significant regression")
+    
+    lines.append("=" * 70)
+    
+    return "\n".join(lines)
@@ -0,0 +1,386 @@
+"""
+Profile Suite Runner for PraisonAI CLI.
+
+Runs a matrix of profiling scenarios and produces aggregated reports.
+Supports cold/warm start, streaming/non-streaming, and multiple iterations.
+"""
+
+import json
+import statistics
+import subprocess
+import sys
+import time
+from dataclasses import dataclass, field
+from datetime import datetime
+from pathlib import Path
+from typing import Any, Dict, List, Optional
+
+from .core import ProfilerConfig, ProfilerResult, QueryProfiler
+
+
+@dataclass
+class ScenarioConfig:
+    """Configuration for a single profiling scenario."""
+    name: str
+    prompt: str
+    model: Optional[str] = None
+    stream: bool = False
+    iterations: int = 3
+    warmup: int = 1
+    show_files: bool = True
+    limit: int = 20
+
+
+@dataclass
+class ScenarioResult:
+    """Result from running a scenario."""
+    name: str
+    config: ScenarioConfig
+    results: List[ProfilerResult] = field(default_factory=list)
+    
+    @property
+    def import_times(self) -> List[float]:
+        return [r.timing.imports_ms for r in self.results]
+    
+    @property
+    def total_times(self) -> List[float]:
+        return [r.timing.total_run_ms for r in self.results]
+    
+    @property
+    def first_token_times(self) -> List[float]:
+        return [r.timing.first_token_ms for r in self.results if r.timing.first_token_ms > 0]
+    
+    def get_stats(self, values: List[float]) -> Dict[str, float]:
+        if not values:
+            return {"mean": 0, "min": 0, "max": 0, "stdev": 0, "median": 0}
+        return {
+            "mean": statistics.mean(values),
+            "min": min(values),
+            "max": max(values),
+            "stdev": statistics.stdev(values) if len(values) > 1 else 0,
+            "median": statistics.median(values),
+        }
+    
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "name": self.name,
+            "config": {
+                "prompt": self.config.prompt[:50] + "..." if len(self.config.prompt) > 50 else self.config.prompt,
+                "model": self.config.model or "default",
+                "stream": self.config.stream,
+                "iterations": self.config.iterations,
+            },
+            "import_time_stats": self.get_stats(self.import_times),
+            "total_time_stats": self.get_stats(self.total_times),
+            "first_token_stats": self.get_stats(self.first_token_times) if self.first_token_times else None,
+            "raw_import_times_ms": self.import_times,
+            "raw_total_times_ms": self.total_times,
+        }
+
+
+@dataclass
+class SuiteResult:
+    """Result from running a full profiling suite."""
+    scenarios: List[ScenarioResult] = field(default_factory=list)
+    startup_cold_ms: float = 0.0
+    startup_warm_ms: float = 0.0
+    import_analysis: List[Dict[str, Any]] = field(default_factory=list)
+    timestamp: str = field(default_factory=lambda: datetime.utcnow().isoformat() + "Z")
+    metadata: Dict[str, Any] = field(default_factory=dict)
+    
+    def to_dict(self) -> Dict[str, Any]:
+        return {
+            "timestamp": self.timestamp,
+            "metadata": self.metadata,
+            "startup": {
+                "cold_ms": self.startup_cold_ms,
+                "warm_ms": self.startup_warm_ms,
+            },
+            "import_analysis": self.import_analysis[:20],
+            "scenarios": [s.to_dict() for s in self.scenarios],
+        }
+
+
+class ProfileSuiteRunner:
+    """
+    Runs a suite of profiling scenarios.
+    
+    Collects comprehensive timing data across multiple runs and scenarios.
+    """
+    
+    DEFAULT_SCENARIOS = [
+        ScenarioConfig(name="simple_non_stream", prompt="hi", stream=False, iterations=3),
+        ScenarioConfig(name="simple_stream", prompt="hi", stream=True, iterations=3),
+        ScenarioConfig(name="medium_non_stream", prompt="Explain what Python is in 2 sentences", stream=False, iterations=2),
+        ScenarioConfig(name="medium_stream", prompt="Explain what Python is in 2 sentences", stream=True, iterations=2),
+    ]
+    
+    def __init__(
+        self,
+        scenarios: Optional[List[ScenarioConfig]] = None,
+        output_dir: Optional[str] = None,
+        verbose: bool = False,
+    ):
+        self.scenarios = scenarios or self.DEFAULT_SCENARIOS
+        self.output_dir = Path(output_dir) if output_dir else Path("/tmp/praisonai_profile_suite")
+        self.verbose = verbose
+    
+    def run(self) -> SuiteResult:
+        """Run the full profiling suite."""
+        result = SuiteResult()
+        result.metadata = self._collect_metadata()
+        
+        # Create output directory
+        self.output_dir.mkdir(parents=True, exist_ok=True)
+        
+        if self.verbose:
+            print("🔬 Starting Profile Suite...")
+            print(f"   Output: {self.output_dir}")
+        
+        # 1. Measure startup times
+        if self.verbose:
+            print("\n📊 Measuring startup times...")
+        result.startup_cold_ms, result.startup_warm_ms = self._measure_startup()
+        if self.verbose:
+            print(f"   Cold: {result.startup_cold_ms:.2f}ms, Warm: {result.startup_warm_ms:.2f}ms")
+        
+        # 2. Analyze imports
+        if self.verbose:
+            print("\n📊 Analyzing imports...")
+        result.import_analysis = self._analyze_imports()
+        if self.verbose and result.import_analysis:
+            print(f"   Top import: {result.import_analysis[0]['module']} ({result.import_analysis[0]['cumulative_ms']:.2f}ms)")
+        
+        # 3. Run scenarios
+        for scenario in self.scenarios:
+            if self.verbose:
+                print(f"\n📊 Running scenario: {scenario.name}")
+            scenario_result = self._run_scenario(scenario)
+            result.scenarios.append(scenario_result)
+            
+            if self.verbose:
+                stats = scenario_result.get_stats(scenario_result.total_times)
+                print(f"   Total time: {stats['mean']:.2f}ms (±{stats['stdev']:.2f}ms)")
+        
+        # 4. Save results
+        self._save_results(result)
+        
+        if self.verbose:
+            print(f"\n✅ Suite complete. Results saved to {self.output_dir}")
+        
+        return result
+    
+    def _collect_metadata(self) -> Dict[str, Any]:
+        """Collect system metadata."""
+        import platform
+        try:
+            from praisonai.version import __version__ as praisonai_version
+        except ImportError:
+            praisonai_version = "unknown"
+        
+        return {
+            "python_version": platform.python_version(),
+            "platform": platform.platform(),
+            "praisonai_version": praisonai_version,
+            "timestamp": datetime.utcnow().isoformat() + "Z",
+        }
+    
+    def _measure_startup(self) -> tuple:
+        """Measure cold and warm startup times."""
+        # Cold start
+        start = time.perf_counter()
+        subprocess.run(
+            [sys.executable, "-c", "import praisonai; import praisonai.cli"],
+            capture_output=True,
+            text=True,
+        )
+        cold_ms = (time.perf_counter() - start) * 1000
+        
+        # Warm start
+        start = time.perf_counter()
+        subprocess.run(
+            [sys.executable, "-c", "import praisonai; import praisonai.cli"],
+            capture_output=True,
+            text=True,
+        )
+        warm_ms = (time.perf_counter() - start) * 1000
+        
+        return cold_ms, warm_ms
+    
+    def _analyze_imports(self) -> List[Dict[str, Any]]:
+        """Analyze import times."""
+        import re
+        
+        try:
+            result = subprocess.run(
+                [sys.executable, "-X", "importtime", "-c", "import praisonaiagents"],
+                capture_output=True,
+                text=True,
+                timeout=60,
+            )
+            
+            import_times = []
+            for line in result.stderr.split('\n'):
+                if 'import time:' in line:
+                    match = re.search(r'import time:\s+(\d+)\s+\|\s+(\d+)\s+\|\s+(.+)', line)
+                    if match:
+                        self_time = int(match.group(1))
+                        cumulative = int(match.group(2))
+                        module = match.group(3).strip()
+                        import_times.append({
+                            "module": module,
+                            "self_us": self_time,
+                            "cumulative_us": cumulative,
+                            "self_ms": self_time / 1000,
+                            "cumulative_ms": cumulative / 1000,
+                        })
+            
+            import_times.sort(key=lambda x: x["cumulative_us"], reverse=True)
+            return import_times[:30]
+            
+        except Exception:
+            return []
+    
+    def _run_scenario(self, scenario: ScenarioConfig) -> ScenarioResult:
+        """Run a single scenario with multiple iterations."""
+        result = ScenarioResult(name=scenario.name, config=scenario)
+        
+        config = ProfilerConfig(
+            stream=scenario.stream,
+            show_files=scenario.show_files,
+            limit=scenario.limit,
+            first_token=scenario.stream,
+        )
+        
+        # Warmup runs (discarded)
+        for _ in range(scenario.warmup):
+            try:
+                profiler = QueryProfiler(config)
+                profiler.profile_query(scenario.prompt, model=scenario.model, stream=scenario.stream)
+            except Exception:
+                pass
+        
+        # Actual runs
+        for i in range(scenario.iterations):
+            try:
+                profiler = QueryProfiler(config)
+                profile_result = profiler.profile_query(
+                    scenario.prompt,
+                    model=scenario.model,
+                    stream=scenario.stream,
+                )
+                result.results.append(profile_result)
+                
+                if self.verbose:
+                    print(f"   Iteration {i+1}: {profile_result.timing.total_run_ms:.2f}ms")
+                    
+            except Exception as e:
+                if self.verbose:
+                    print(f"   Iteration {i+1}: ERROR - {e}")
+        
+        return result
+    
+    def _save_results(self, result: SuiteResult):
+        """Save results to files."""
+        # Save JSON summary
+        json_path = self.output_dir / "suite_results.json"
+        with open(json_path, 'w') as f:
+            json.dump(result.to_dict(), f, indent=2, default=str)
+        
+        # Save human-readable report
+        report_path = self.output_dir / "suite_report.txt"
+        with open(report_path, 'w') as f:
+            f.write(self._format_report(result))
+    
+    def _format_report(self, result: SuiteResult) -> str:
+        """Format a human-readable report."""
+        lines = []
+        lines.append("=" * 70)
+        lines.append("PraisonAI Profile Suite Report")
+        lines.append("=" * 70)
+        lines.append("")
+        
+        # Metadata
+        lines.append("## System Information")
+        lines.append(f"  Timestamp:        {result.timestamp}")
+        lines.append(f"  Python Version:   {result.metadata.get('python_version', 'N/A')}")
+        lines.append(f"  Platform:         {result.metadata.get('platform', 'N/A')}")
+        lines.append(f"  PraisonAI:        {result.metadata.get('praisonai_version', 'N/A')}")
+        lines.append("")
+        
+        # Startup times
+        lines.append("## Startup Times")
+        lines.append(f"  Cold Start:       {result.startup_cold_ms:>10.2f} ms")
+        lines.append(f"  Warm Start:       {result.startup_warm_ms:>10.2f} ms")
+        lines.append("")
+        
+        # Import analysis
+        if result.import_analysis:
+            lines.append("## Top Import Times")
+            lines.append("-" * 70)
+            lines.append(f"{'Module':<45} {'Self (ms)':>10} {'Cumul (ms)':>12}")
+            lines.append("-" * 70)
+            for imp in result.import_analysis[:15]:
+                module = imp["module"]
+                if len(module) > 43:
+                    module = "..." + module[-40:]
+                lines.append(f"{module:<45} {imp['self_ms']:>10.2f} {imp['cumulative_ms']:>12.2f}")
+            lines.append("")
+        
+        # Scenario results
+        lines.append("## Scenario Results")
+        lines.append("-" * 70)
+        lines.append(f"{'Scenario':<25} {'Stream':>8} {'Iters':>6} {'Mean (ms)':>12} {'StdDev':>10}")
+        lines.append("-" * 70)
+        
+        for scenario in result.scenarios:
+            stats = scenario.get_stats(scenario.total_times)
+            stream_str = "Yes" if scenario.config.stream else "No"
+            lines.append(
+                f"{scenario.name:<25} {stream_str:>8} {len(scenario.results):>6} "
+                f"{stats['mean']:>12.2f} {stats['stdev']:>10.2f}"
+            )
+        lines.append("")
+        
+        # Summary statistics
+        all_import_times = []
+        all_total_times = []
+        for scenario in result.scenarios:
+            all_import_times.extend(scenario.import_times)
+            all_total_times.extend(scenario.total_times)
+        
+        if all_import_times:
+            lines.append("## Overall Statistics")
+            lines.append(f"  Import Time (mean):   {statistics.mean(all_import_times):>10.2f} ms")
+            lines.append(f"  Total Time (mean):    {statistics.mean(all_total_times):>10.2f} ms")
+            if len(all_total_times) > 1:
+                lines.append(f"  Total Time (stdev):   {statistics.stdev(all_total_times):>10.2f} ms")
+            lines.append("")
+        
+        lines.append("=" * 70)
+        
+        return "\n".join(lines)
+
+
+def run_profile_suite(
+    output_dir: Optional[str] = None,
+    scenarios: Optional[List[ScenarioConfig]] = None,
+    verbose: bool = True,
+) -> SuiteResult:
+    """
+    Run a profile suite and return results.
+    
+    Args:
+        output_dir: Directory to save results
+        scenarios: Custom scenarios to run
+        verbose: Print progress
+        
+    Returns:
+        SuiteResult with all profiling data
+    """
+    runner = ProfileSuiteRunner(
+        scenarios=scenarios,
+        output_dir=output_dir,
+        verbose=verbose,
+    )
+    return runner.run()
@@ -29,6 +29,10 @@ class QueueManager:
    - Event handling
    """
    
+    # Runtime tools registry - stores tools that can't be serialized to JSON
+    # Key: run_id or "default", Value: list of tool callables
+    _tools_registry: Dict[str, list] = {}
+    
    def __init__(
        self,
        config: Optional[QueueConfig] = None,
@@ -36,6 +40,7 @@ class QueueManager:
        on_complete: Optional[Callable[[str, QueuedRun], Coroutine[Any, Any, None]]] = None,
        on_error: Optional[Callable[[str, Exception], Coroutine[Any, Any, None]]] = None,
        on_event: Optional[Callable[[QueueEvent], Coroutine[Any, Any, None]]] = None,
+        default_tools: Optional[list] = None,
    ):
        """
        Initialize queue manager.
@@ -46,6 +51,7 @@ class QueueManager:
            on_complete: Callback when run completes.
            on_error: Callback when run fails.
            on_event: Callback for queue events.
+            default_tools: Default tools to use for all runs (stored in runtime registry).
        """
        self.config = config or QueueConfig()
        
@@ -70,6 +76,10 @@ class QueueManager:
        
        # Session tracking
        self._current_session_id: Optional[str] = None
+        
+        # Store default tools in registry
+        if default_tools:
+            QueueManager._tools_registry["default"] = default_tools
    
    async def start(self, recover: bool = True) -> None:
        """
@@ -157,6 +167,7 @@ class QueueManager:
        workspace: Optional[str] = None,
        config: Optional[Dict[str, Any]] = None,
        run_id: Optional[str] = None,
+        chat_history: Optional[List[Dict[str, str]]] = None,
    ) -> str:
        """
        Submit a new run to the queue.
@@ -169,6 +180,7 @@ class QueueManager:
            workspace: Workspace path.
            config: Additional configuration.
            run_id: Optional specific run ID.
+            chat_history: Previous chat messages for context continuity.
            
        Returns:
            The run ID.
@@ -181,6 +193,7 @@ class QueueManager:
            workspace=workspace or os.getcwd(),
            config=config or {},
            max_retries=self.config.default_max_retries,
+            chat_history=chat_history or [],
        )
        
        if run_id:
@@ -225,6 +238,10 @@ class QueueManager:
        """Resume a paused run."""
        return await self.scheduler.resume(run_id)
    
+    async def update_input(self, run_id: str, new_input: str) -> bool:
+        """Update the input content of a queued run."""
+        return await self.scheduler.update_input(run_id, new_input)
+    
    def get_run(self, run_id: str) -> Optional[QueuedRun]:
        """Get a run by ID."""
        return self.scheduler.get_run(run_id)
@@ -393,3 +410,22 @@ class QueueManager:
    def running_count(self) -> int:
        """Number of running runs."""
        return self.scheduler.running_count
+    
+    @classmethod
+    def get_tools_for_run(cls, run_id: str) -> list:
+        """
+        Get tools for a specific run from the registry.
+        
+        Falls back to default tools if no run-specific tools are registered.
+        """
+        return cls._tools_registry.get(run_id) or cls._tools_registry.get("default", [])
+    
+    @classmethod
+    def register_tools(cls, run_id: str, tools: list) -> None:
+        """Register tools for a specific run."""
+        cls._tools_registry[run_id] = tools
+    
+    @classmethod
+    def unregister_tools(cls, run_id: str) -> None:
+        """Unregister tools for a specific run."""
+        cls._tools_registry.pop(run_id, None)
@@ -85,6 +85,9 @@ class QueuedRun:
    # Configuration
    config: Dict[str, Any] = field(default_factory=dict)
    
+    # Chat history for session continuity
+    chat_history: List[Dict[str, str]] = field(default_factory=list)
+    
    # Streaming state
    output_chunks: List[str] = field(default_factory=list)
    
@@ -302,6 +302,40 @@ class QueueScheduler:
        """Clear a cancellation token."""
        self._cancel_tokens.discard(run_id)
    
+    async def update_input(self, run_id: str, new_input: str) -> bool:
+        """
+        Update the input content of a queued run.
+        
+        Only works for runs in QUEUED state.
+        
+        Args:
+            run_id: The run ID to update.
+            new_input: The new input content.
+            
+        Returns:
+            True if updated, False if not found or not editable.
+        """
+        async with self._lock:
+            run = self._all_runs.get(run_id)
+            if run is None:
+                return False
+            
+            if run.state != RunState.QUEUED:
+                logger.warning(f"Cannot edit run {run_id} in state {run.state}")
+                return False
+            
+            run.input_content = new_input
+            
+            logger.debug(f"Updated input for run {run_id}")
+            
+            self._emit_event(QueueEvent(
+                event_type="run_updated",
+                run_id=run_id,
+                data={"new_input": new_input[:50]}
+            ))
+            
+            return True
+    
    async def retry(self, run_id: str) -> Optional[str]:
        """
        Retry a failed run.
@@ -210,20 +210,31 @@ class WorkerPool:
        try:
            # Import agent lazily to avoid circular imports
            from praisonaiagents import Agent
+            from .manager import QueueManager
            
            # Get agent configuration
            agent_config = run.config.get("agent_config", {})
            agent_name = run.agent_name or agent_config.get("name", "Assistant")
            
-            # Create agent
+            # Get tools from runtime registry (not from config, as functions can't be serialized)
+            tools = QueueManager.get_tools_for_run(run.run_id)
+            
+            # Create agent with session_id for history persistence
+            # Note: Agent uses 'llm' parameter, not 'model'
            agent = Agent(
                name=agent_name,
                instructions=agent_config.get("instructions", "You are a helpful assistant."),
-                model=agent_config.get("model"),
-                tools=agent_config.get("tools", []),
+                llm=agent_config.get("model"),
+                tools=tools,
                verbose=agent_config.get("verbose", False),
+                session_id=agent_config.get("session_id") or run.session_id,  # Use session for history
            )
            
+            # Inject chat history from session store for context continuity
+            if run.chat_history:
+                agent.chat_history = list(run.chat_history)
+                logger.debug(f"Injected {len(run.chat_history)} messages from session history")
+            
            # Check for cancellation before starting
            if self.scheduler.is_cancelled(run.run_id):
                logger.debug(f"Run {run.run_id} was cancelled before execution")
@@ -13,12 +13,12 @@ from typing import Any, Dict, List, Optional
 try:
    from textual.app import App, ComposeResult
    from textual.binding import Binding
-    from textual.screen import Screen
-    from textual import events
+    from textual import work
    TEXTUAL_AVAILABLE = True
 except ImportError:
    TEXTUAL_AVAILABLE = False
    App = object
+    work = None

 logger = logging.getLogger(__name__)

@@ -28,10 +28,12 @@ if TEXTUAL_AVAILABLE:
    from .screens.queue import QueueScreen
    from .screens.settings import SettingsScreen
    from .screens.session import SessionScreen
+    from .screens.help import HelpScreen
    from .widgets.chat import ChatMessage
    from .widgets.status import StatusInfo
    from .widgets.queue_panel import QueueItem
    from ..queue import QueueManager, QueueConfig, QueuedRun, RunPriority
+    from .session_store import SessionStore, get_session_store

    class TUIApp(App):
        """
@@ -54,9 +56,12 @@ if TEXTUAL_AVAILABLE:
        }
        """
        
+        # Safe default bindings - Ctrl keys only at app level for emergency exit
+        # MainScreen handles all other bindings with safe defaults
        BINDINGS = [
-            Binding("ctrl+q", "quit", "Quit", show=True, priority=True),
-            Binding("ctrl+c", "cancel", "Cancel", show=True),
+            # Emergency quit - kept at app level for reliability
+            Binding("ctrl+q", "quit", "Quit", show=False, priority=True),
+            Binding("ctrl+c", "cancel", "Cancel", show=False),
        ]
        
        SCREENS = {
@@ -64,6 +69,7 @@ if TEXTUAL_AVAILABLE:
            "queue": QueueScreen,
            "settings": SettingsScreen,
            "sessions": SessionScreen,
+            "help": HelpScreen,
        }
        
        def __init__(
@@ -73,6 +79,8 @@ if TEXTUAL_AVAILABLE:
            model: Optional[str] = None,
            agent_config: Optional[Dict[str, Any]] = None,
            queue_config: Optional[QueueConfig] = None,
+            enable_acp: bool = True,
+            enable_lsp: bool = True,
        ):
            super().__init__()
            
@@ -80,11 +88,20 @@ if TEXTUAL_AVAILABLE:
            self.session_id = session_id or str(uuid.uuid4())[:8]
            self.model = model or os.environ.get("OPENAI_MODEL", "gpt-4o-mini")
            self.agent_config = agent_config or {}
+            self.enable_acp = enable_acp
+            self.enable_lsp = enable_lsp
+            
+            # Load default interactive tools (ACP + LSP) if not provided
+            if "tools" not in self.agent_config or not self.agent_config["tools"]:
+                self.agent_config["tools"] = self._load_default_tools()
            
            # Queue manager
            self.queue_config = queue_config or QueueConfig()
            self.queue_manager: Optional[QueueManager] = None
            
+            # Session store for chat history persistence
+            self.session_store = get_session_store()
+            
            # State
            self._current_run_id: Optional[str] = None
            self._streaming_content: str = ""
@@ -95,6 +112,43 @@ if TEXTUAL_AVAILABLE:
            self._command_handlers: Dict[str, callable] = {}
            self._setup_commands()
        
+        def _load_default_tools(self) -> list:
+            """
+            Load default interactive tools (ACP + LSP) for TUI.
+            
+            Uses the canonical interactive_tools provider.
+            """
+            try:
+                from ..interactive_tools import get_interactive_tools, ToolConfig
+                
+                # Build disable list based on flags
+                disable = []
+                if not self.enable_acp:
+                    disable.append('acp')
+                if not self.enable_lsp:
+                    disable.append('lsp')
+                
+                config = ToolConfig(
+                    workspace=self.workspace,
+                    enable_acp=self.enable_acp,
+                    enable_lsp=self.enable_lsp,
+                )
+                
+                tools = get_interactive_tools(
+                    config=config,
+                    disable=disable if disable else None,
+                )
+                
+                logger.info(f"TUI loaded {len(tools)} default tools (ACP: {self.enable_acp}, LSP: {self.enable_lsp})")
+                return tools
+                
+            except ImportError as e:
+                logger.warning(f"Could not load interactive tools: {e}")
+                return []
+            except Exception as e:
+                logger.error(f"Error loading default tools: {e}")
+                return []
+        
        def _setup_commands(self) -> None:
            """Setup slash command handlers."""
            self._command_handlers = {
@@ -109,16 +163,25 @@ if TEXTUAL_AVAILABLE:
                "cost": self._cmd_cost,
                "exit": self._cmd_exit,
                "quit": self._cmd_exit,
+                # Additional commands
+                "tokens": self._cmd_tokens,
+                "plan": self._cmd_plan,
+                "map": self._cmd_map,
+                "undo": self._cmd_undo,
+                "diff": self._cmd_diff,
+                "commit": self._cmd_commit,
+                "tools": self._cmd_tools,
            }
        
        async def on_mount(self) -> None:
            """Handle app mount."""
-            # Initialize queue manager
+            # Initialize queue manager with default tools (stored in runtime registry)
            self.queue_manager = QueueManager(
                config=self.queue_config,
                on_output=self._handle_output,
                on_complete=self._handle_complete,
                on_error=self._handle_error,
+                default_tools=self.agent_config.get("tools", []),
            )
            
            await self.queue_manager.start(recover=True)
@@ -127,6 +190,9 @@ if TEXTUAL_AVAILABLE:
            # Push main screen
            await self.push_screen("main")
            
+            # Update tools panel with available tools
+            self._update_tools_panel()
+            
            # Update status
            self._update_status()
        
@@ -141,23 +207,45 @@ if TEXTUAL_AVAILABLE:
        
        # Event handlers
        
-        async def on_main_screen_message_submitted(
+        def on_main_screen_message_submitted(
            self, event: MainScreen.MessageSubmitted
        ) -> None:
-            """Handle message submission from main screen."""
-            content = event.content
+            """Handle message submission - dispatch to worker for non-blocking execution."""
+            # Use run_worker to avoid blocking the UI
+            self.run_worker(
+                self._process_message_submission(event.content),
+                exclusive=False,  # Allow multiple messages to queue
+                exit_on_error=False,
+            )
+        
+        @work(exclusive=False, exit_on_error=False)
+        async def _process_message_submission(self, content: str) -> None:
+            """Process message submission in background worker.
            
-            # Add user message to chat
+            Using @work decorator ensures the UI remains responsive.
+            User can type while agent is processing.
+            """
+            # Add user message to chat display
            main_screen = self.screen
            if isinstance(main_screen, MainScreen):
                await main_screen.add_user_message(content)
                main_screen.set_processing(True)
            
-            # Submit to queue
+            # Add user message to session store for history persistence
+            self.session_store.add_user_message(self.session_id, content)
+            
+            # Get chat history for context continuity
+            chat_history = self.session_store.get_chat_history(self.session_id, max_messages=50)
+            
+            # Submit to queue with chat history
+            # Note: tools are NOT included in config (they can't be JSON serialized)
+            # Instead, they are stored in QueueManager._tools_registry and retrieved by worker
            try:
                run_id = await self.queue_manager.submit(
                    input_content=content,
                    agent_name=self.agent_config.get("name", "Assistant"),
+                    session_id=self.session_id,  # Pass session_id for history
+                    chat_history=chat_history,  # Pass previous messages for context
                    config={
                        "agent_config": {
                            "name": self.agent_config.get("name", "Assistant"),
@@ -166,8 +254,8 @@ if TEXTUAL_AVAILABLE:
                                "You are a helpful AI assistant."
                            ),
                            "model": self.model,
-                            "tools": self.agent_config.get("tools", []),
                            "verbose": False,
+                            "session_id": self.session_id,  # Pass to agent for history
                        }
                    }
                )
@@ -175,6 +263,9 @@ if TEXTUAL_AVAILABLE:
                self._current_run_id = run_id
                self._streaming_content = ""
                
+                # IMMEDIATELY update queue panel - increment the moment message is queued
+                self._update_queue_panel_live()
+                
                # Add streaming placeholder
                if isinstance(main_screen, MainScreen):
                    await main_screen.add_assistant_message(
@@ -194,6 +285,7 @@ if TEXTUAL_AVAILABLE:
                    )
            
            self._update_status()
+            self._update_queue_panel_live()
        
        async def on_main_screen_command_executed(
            self, event: MainScreen.CommandExecuted
@@ -278,6 +370,22 @@ if TEXTUAL_AVAILABLE:
            self.queue_manager.set_session(self.session_id)
            self._update_status()
        
+        async def on_queue_panel_widget_edit_requested(
+            self, event
+        ) -> None:
+            """Handle queue item edit request."""
+            from .widgets.queue_panel import QueuePanelWidget
+            if isinstance(event, QueuePanelWidget.EditRequested):
+                success = await self.queue_manager.update_input(event.run_id, event.new_content)
+                if success:
+                    self._update_queue_panel_live()
+                    main_screen = self.screen
+                    if isinstance(main_screen, MainScreen):
+                        await main_screen.add_assistant_message(
+                            content=f"Updated queue item {event.run_id[:8]}",
+                            agent_name="System",
+                        )
+        
        # Queue callbacks
        
        async def _handle_output(self, run_id: str, chunk: str) -> None:
@@ -293,14 +401,17 @@ if TEXTUAL_AVAILABLE:
        
        async def _handle_complete(self, run_id: str, run: QueuedRun) -> None:
            """Handle run completion."""
+            output_content = run.output_content or self._streaming_content
+            
            main_screen = self.screen
            if isinstance(main_screen, MainScreen):
-                await main_screen.complete_streaming(
-                    run_id, 
-                    run.output_content or self._streaming_content
-                )
+                await main_screen.complete_streaming(run_id, output_content)
                main_screen.set_processing(False)
            
+            # Store assistant response in session store for history persistence
+            if output_content and run.session_id:
+                self.session_store.add_assistant_message(run.session_id, output_content, run_id)
+            
            # Update metrics
            if run.metrics:
                self._total_tokens += run.metrics.get("tokens", 0)
@@ -311,6 +422,7 @@ if TEXTUAL_AVAILABLE:
                self._streaming_content = ""
            
            self._update_status()
+            self._update_queue_panel_live()
        
        async def _handle_error(self, run_id: str, error: Exception) -> None:
            """Handle run error."""
@@ -324,6 +436,7 @@ if TEXTUAL_AVAILABLE:
                self._streaming_content = ""
            
            self._update_status()
+            self._update_queue_panel_live()
        
        # Command handlers
        
@@ -343,18 +456,23 @@ if TEXTUAL_AVAILABLE:
 - `/cost` - Show cost summary
 - `/exit` or `/quit` - Exit TUI

-**Keyboard Shortcuts:**
+**Keyboard Shortcuts (Safe Defaults):**

- `Ctrl+Enter` - Send message
- `Ctrl+C` - Cancel current operation
- `Ctrl+Q` - Quit
- `Ctrl+L` - Clear screen
- `F1` - Help
- `F2` - Toggle queue panel
- `F3` - Settings
- `F5` - Clear chat
- `Tab` - Cycle focus
+- `Enter` - Send message
+- `Shift+Enter` - New line
+- `q` - Quit (when not typing)
+- `:` - Command mode
+- `?` - Help
 - `Esc` - Cancel/close
+
+**Command Mode:**
+
+Press `:` then type a command:
+- `:quit` or `:q` - Quit
+- `:clear` or `:cl` - Clear chat
+- `:help` or `:h` - Help
+- `:tools` or `:t` - Toggle tools
+- `:queue` or `:qu` - Toggle queue
 """
            main_screen = self.screen
            if isinstance(main_screen, MainScreen):
@@ -431,6 +549,8 @@ if TEXTUAL_AVAILABLE:
        async def _cmd_sessions(self, args: str) -> None:
            """Show sessions."""
            await self.push_screen("sessions")
+            # Load and display sessions
+            await self._update_sessions_screen()
        
        async def _cmd_cost(self, args: str) -> None:
            """Show cost summary."""
@@ -449,6 +569,115 @@ if TEXTUAL_AVAILABLE:
            """Exit TUI."""
            self.exit()
        
+        async def _cmd_tokens(self, args: str) -> None:
+            """Show token usage breakdown."""
+            msg = f"""
+**Token Usage:**
+
+- Total Tokens: {self._total_tokens:,}
+- Estimated Cost: ${self._total_cost:.4f}
+- Session: {self.session_id}
+"""
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                await main_screen.add_assistant_message(content=msg, agent_name="System")
+        
+        async def _cmd_plan(self, args: str) -> None:
+            """Show or create a plan."""
+            msg = """
+**Plan Command:**
+
+The /plan command is used to create or view task plans.
+Currently, plans are managed through the agent's task execution.
+
+Usage:
+- `/plan` - View current plan
+- `/plan <task>` - Create a plan for a task
+"""
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                await main_screen.add_assistant_message(content=msg, agent_name="System")
+        
+        async def _cmd_map(self, args: str) -> None:
+            """Show repository map."""
+            msg = """
+**Repository Map:**
+
+The /map command shows a map of the repository structure.
+Use the file tools to explore the codebase.
+
+Available tools:
+- `acp_list_files` - List files in a directory
+- `acp_read_file` - Read file contents
+"""
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                await main_screen.add_assistant_message(content=msg, agent_name="System")
+        
+        async def _cmd_undo(self, args: str) -> None:
+            """Undo last change."""
+            msg = """
+**Undo:**
+
+The /undo command reverts the last file change.
+Use git commands or the agent to manage file changes.
+
+Tip: Ask the agent to "undo the last change" or use git.
+"""
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                await main_screen.add_assistant_message(content=msg, agent_name="System")
+        
+        async def _cmd_diff(self, args: str) -> None:
+            """Show diff of changes."""
+            msg = """
+**Diff:**
+
+The /diff command shows changes made to files.
+Use git diff or ask the agent to show changes.
+
+Tip: Ask the agent "show me the diff" or use `git diff`.
+"""
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                await main_screen.add_assistant_message(content=msg, agent_name="System")
+        
+        async def _cmd_commit(self, args: str) -> None:
+            """Commit changes."""
+            msg = """
+**Commit:**
+
+The /commit command creates a git commit.
+Use git commands or ask the agent to commit changes.
+
+Tip: Ask the agent to "commit the changes with message X".
+"""
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                await main_screen.add_assistant_message(content=msg, agent_name="System")
+        
+        async def _cmd_tools(self, args: str) -> None:
+            """Show available tools."""
+            tools = self.agent_config.get("tools", [])
+            tool_names = []
+            for tool in tools:
+                if hasattr(tool, '__name__'):
+                    tool_names.append(tool.__name__)
+                elif hasattr(tool, 'name'):
+                    tool_names.append(tool.name)
+                else:
+                    tool_names.append(str(tool)[:30])
+            
+            msg = f"""
+**Available Tools ({len(tools)}):**
+
+{chr(10).join(f'- {name}' for name in tool_names[:20])}
+{"..." if len(tool_names) > 20 else ""}
+"""
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                await main_screen.add_assistant_message(content=msg, agent_name="System")
+        
        # Helper methods
        
        def _update_status(self) -> None:
@@ -465,6 +694,19 @@ if TEXTUAL_AVAILABLE:
                    is_processing=self._current_run_id is not None,
                ))
        
+        def _update_queue_panel_live(self) -> None:
+            """Update queue panel on MainScreen with live counts."""
+            from .widgets.queue_panel import QueuePanelWidget
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                try:
+                    queue_panel = main_screen.query_one("#queue-panel", QueuePanelWidget)
+                    # Update counts immediately
+                    queue_panel.queued_count = self.queue_manager.queued_count if self.queue_manager else 0
+                    queue_panel.running_count = self.queue_manager.running_count if self.queue_manager else 0
+                except Exception:
+                    pass
+        
        def _update_queue_screen(self) -> None:
            """Update queue screen if visible."""
            if isinstance(self.screen, QueueScreen):
@@ -482,6 +724,41 @@ if TEXTUAL_AVAILABLE:
                    for r in runs
                ])
        
+        async def _update_sessions_screen(self) -> None:
+            """Update sessions screen with session list."""
+            if isinstance(self.screen, SessionScreen):
+                # Get sessions from queue manager or session store
+                sessions = []
+                if self.queue_manager:
+                    try:
+                        # Try to get sessions from the persistence layer
+                        from ..queue.persistence import QueuePersistence
+                        persistence = QueuePersistence()
+                        sessions = persistence.list_sessions(limit=50)
+                    except Exception as e:
+                        logger.debug(f"Could not load sessions: {e}")
+                        # Fallback: show current session only
+                        sessions = [{
+                            "session_id": self.session_id,
+                            "created_at": "-",
+                            "updated_at": "-",
+                            "run_count": self.queue_manager.queued_count + self.queue_manager.running_count,
+                        }]
+                
+                self.screen.update_sessions(sessions)
+        
+        def _update_tools_panel(self) -> None:
+            """Update tools panel with available tools."""
+            from .widgets.tool_panel import ToolPanelWidget
+            main_screen = self.screen
+            if isinstance(main_screen, MainScreen):
+                try:
+                    tool_panel = main_screen.query_one("#tool-panel", ToolPanelWidget)
+                    tools = self.agent_config.get("tools", [])
+                    tool_panel.set_available_tools(tools)
+                except Exception:
+                    pass
+        
        def action_quit(self) -> None:
            """Quit the application."""
            self.exit()
@@ -505,6 +782,8 @@ def run_tui(
    session_id: Optional[str] = None,
    model: Optional[str] = None,
    agent_config: Optional[Dict[str, Any]] = None,
+    enable_acp: bool = True,
+    enable_lsp: bool = True,
 ) -> None:
    """
    Run the PraisonAI TUI.
@@ -514,6 +793,8 @@ def run_tui(
        session_id: Session ID to resume.
        model: Default model.
        agent_config: Agent configuration.
+        enable_acp: Enable ACP tools (default: True).
+        enable_lsp: Enable LSP tools (default: True).
    """
    if not TEXTUAL_AVAILABLE:
        raise ImportError(
@@ -525,5 +806,7 @@ def run_tui(
        session_id=session_id,
        model=model,
        agent_config=agent_config,
+        enable_acp=enable_acp,
+        enable_lsp=enable_lsp,
    )
    app.run()
@@ -56,8 +56,16 @@ def create_tui_app():
            None, "--agent", "-a",
            help="Agent config file (YAML)"
        ),
+        no_acp: bool = typer.Option(
+            False, "--no-acp",
+            help="Disable ACP tools (file operations with plan/approve/apply)"
+        ),
+        no_lsp: bool = typer.Option(
+            False, "--no-lsp",
+            help="Disable LSP tools (code intelligence: symbols, definitions)"
+        ),
    ):
-        """Launch the interactive TUI."""
+        """Launch the interactive TUI with ACP + LSP tools enabled by default."""
        try:
            from .app import run_tui
        except ImportError:
@@ -79,6 +87,8 @@ def create_tui_app():
            session_id=session,
            model=model,
            agent_config=agent_config,
+            enable_acp=not no_acp,
+            enable_lsp=not no_lsp,
        )
    
    return app
@@ -0,0 +1,150 @@
+"""
+TUI Configuration for PraisonAI.
+
+Handles keyboard shortcuts and display settings with safe defaults.
+"""
+
+import os
+from dataclasses import dataclass, field
+from typing import Dict, List, Optional
+
+
+@dataclass
+class TUIConfig:
+    """
+    Configuration for TUI behavior.
+    
+    Safe defaults:
+    - Ctrl keys disabled by default (conflict with IDEs/terminals)
+    - Function keys disabled by default (conflict with IDEs/OS)
+    - Single-character shortcuts enabled by default
+    - Command mode (:) enabled by default
+    """
+    
+    # Keyboard configuration
+    enable_ctrl_keys: bool = False  # Ctrl+Q, Ctrl+L, etc.
+    enable_fn_keys: bool = False    # F1-F12
+    enable_single_char: bool = True # q, l, etc. (when not in input)
+    enable_command_mode: bool = True # :quit, :clear, etc.
+    leader_key: str = ":"           # Command mode prefix
+    
+    # Display configuration
+    auto_scroll: bool = True
+    show_tools_panel: bool = True
+    show_queue_panel: bool = True
+    
+    @classmethod
+    def from_env(cls) -> "TUIConfig":
+        """Create config from environment variables."""
+        config = cls()
+        
+        # Check env vars for enabling Ctrl keys (opt-in)
+        if os.environ.get("PRAISONAI_TUI_CTRL_KEYS", "").lower() in ("1", "true", "yes"):
+            config.enable_ctrl_keys = True
+        
+        # Check env vars for enabling Function keys (opt-in)
+        if os.environ.get("PRAISONAI_TUI_FN_KEYS", "").lower() in ("1", "true", "yes"):
+            config.enable_fn_keys = True
+        
+        # Check for disabling single-char shortcuts
+        if os.environ.get("PRAISONAI_TUI_NO_SINGLE_CHAR", "").lower() in ("1", "true", "yes"):
+            config.enable_single_char = False
+        
+        # Check for custom leader key
+        leader = os.environ.get("PRAISONAI_TUI_LEADER", "")
+        if leader and len(leader) == 1:
+            config.leader_key = leader
+        
+        return config
+    
+    @classmethod
+    def from_cli_args(
+        cls,
+        enable_ctrl_keys: bool = False,
+        enable_fn_keys: bool = False,
+        leader_key: Optional[str] = None,
+    ) -> "TUIConfig":
+        """Create config from CLI arguments, merged with env."""
+        config = cls.from_env()
+        
+        # CLI args override env
+        if enable_ctrl_keys:
+            config.enable_ctrl_keys = True
+        if enable_fn_keys:
+            config.enable_fn_keys = True
+        if leader_key and len(leader_key) == 1:
+            config.leader_key = leader_key
+        
+        return config
+
+
+# Command mode commands
+COMMANDS = {
+    "quit": {"aliases": ["q", "exit"], "description": "Quit the TUI"},
+    "clear": {"aliases": ["cl"], "description": "Clear the chat history"},
+    "help": {"aliases": ["h", "?"], "description": "Show help"},
+    "tools": {"aliases": ["t"], "description": "Toggle tools panel"},
+    "queue": {"aliases": ["qu"], "description": "Toggle queue panel"},
+    "settings": {"aliases": ["set"], "description": "Open settings"},
+    "cancel": {"aliases": ["c"], "description": "Cancel current operation"},
+    "sessions": {"aliases": ["sess"], "description": "Manage sessions"},
+}
+
+
+def get_command(input_str: str) -> Optional[str]:
+    """
+    Parse command input and return the canonical command name.
+    
+    Args:
+        input_str: The command string (without leader key)
+        
+    Returns:
+        Canonical command name or None if not found
+    """
+    input_str = input_str.strip().lower()
+    
+    # Check direct match
+    if input_str in COMMANDS:
+        return input_str
+    
+    # Check aliases
+    for cmd, info in COMMANDS.items():
+        if input_str in info.get("aliases", []):
+            return cmd
+    
+    return None
+
+
+def get_help_text() -> str:
+    """Get formatted help text for all commands."""
+    lines = [
+        "PraisonAI TUI Commands",
+        "=" * 40,
+        "",
+        "Command Mode (press : to enter):",
+    ]
+    
+    for cmd, info in COMMANDS.items():
+        aliases = ", ".join(info.get("aliases", []))
+        desc = info.get("description", "")
+        if aliases:
+            lines.append(f"  :{cmd} (or :{aliases}) - {desc}")
+        else:
+            lines.append(f"  :{cmd} - {desc}")
+    
+    lines.extend([
+        "",
+        "Single-key shortcuts (when not typing):",
+        "  q - Quit",
+        "  l - Clear screen",
+        "  ? - Help",
+        "  / - Search/filter",
+        "",
+        "Input shortcuts:",
+        "  Enter - Send message",
+        "  Shift+Enter - New line",
+        "  Escape - Cancel",
+        "  Up/Down - History navigation",
+    ])
+    
+    return "\n".join(lines)
@@ -0,0 +1,157 @@
+"""
+Help Screen for PraisonAI TUI.
+
+Displays keyboard shortcuts and usage information.
+"""
+
+from typing import Optional
+
+try:
+    from textual.screen import Screen
+    from textual.containers import Vertical, Container
+    from textual.widgets import Static, Footer, Button
+    from textual.binding import Binding
+    TEXTUAL_AVAILABLE = True
+except ImportError:
+    TEXTUAL_AVAILABLE = False
+    Screen = object
+
+
+if TEXTUAL_AVAILABLE:
+    from rich.text import Text
+    from rich.panel import Panel
+    from rich.table import Table
+
+    class HelpScreen(Screen):
+        """
+        Help screen showing keyboard shortcuts and usage.
+        """
+        
+        BINDINGS = [
+            Binding("escape", "dismiss", "Close", show=True),
+            Binding("q", "dismiss", "Close", show=True),
+        ]
+        
+        DEFAULT_CSS = """
+        HelpScreen {
+            align: center middle;
+        }
+        
+        HelpScreen #help-container {
+            width: 80;
+            height: auto;
+            max-height: 90%;
+            background: $surface;
+            border: solid $primary;
+            padding: 1 2;
+        }
+        
+        HelpScreen #help-title {
+            text-align: center;
+            text-style: bold;
+            margin-bottom: 1;
+        }
+        
+        HelpScreen #close-button {
+            margin-top: 1;
+            width: 100%;
+        }
+        """
+        
+        def compose(self):
+            """Compose the help screen."""
+            with Container(id="help-container"):
+                yield Static("PraisonAI TUI Help", id="help-title")
+                yield Static(self._build_help_content())
+                yield Button("Close (Esc)", id="close-button", variant="primary")
+        
+        def _build_help_content(self) -> Text:
+            """Build the help content."""
+            content = Text()
+            
+            # Safe keyboard shortcuts section
+            content.append("\n📋 Keyboard Shortcuts (Safe Defaults)\n", style="bold cyan")
+            content.append("─" * 40 + "\n", style="dim")
+            
+            shortcuts = [
+                ("\\", "Show command popup"),
+                ("@path", "Reference file (type freely)"),
+                ("/", "Show command discovery"),
+                ("Enter", "Send message"),
+                ("Shift+Enter", "New line in message"),
+                ("q", "Quit (when not typing)"),
+                ("?", "Show this help"),
+                (":", "Enter command mode"),
+                ("Escape", "Cancel/Close/Dismiss"),
+                ("Tab", "Focus next element"),
+                ("Up/Down", "Navigate history"),
+            ]
+            
+            for key, desc in shortcuts:
+                content.append(f"  {key:<15}", style="yellow")
+                content.append(f" {desc}\n")
+            
+            # Command mode section
+            content.append("\n⌨️  Command Mode (: prefix)\n", style="bold cyan")
+            content.append("─" * 40 + "\n", style="dim")
+            
+            colon_cmds = [
+                (":quit, :q", "Quit application"),
+                (":clear, :cl", "Clear chat history"),
+                (":help, :h", "Show help"),
+                (":tools, :t", "Toggle tools panel"),
+                (":queue, :qu", "Toggle queue panel"),
+                (":settings", "Open settings"),
+                (":cancel", "Cancel current operation"),
+            ]
+            
+            for cmd, desc in colon_cmds:
+                content.append(f"  {cmd:<15}", style="green")
+                content.append(f" {desc}\n")
+            
+            # Slash commands section
+            content.append("\n💬 Slash Commands (/ prefix)\n", style="bold cyan")
+            content.append("─" * 40 + "\n", style="dim")
+            
+            commands = [
+                ("/help", "Show help"),
+                ("/clear", "Clear chat"),
+                ("/model", "Show/change model"),
+                ("/queue", "Show queue status"),
+                ("/cancel", "Cancel current run"),
+                ("/cost", "Show cost summary"),
+                ("/settings", "Open settings"),
+                ("/sessions", "Manage sessions"),
+                ("/exit", "Exit TUI"),
+            ]
+            
+            for cmd, desc in commands:
+                content.append(f"  {cmd:<15}", style="green")
+                content.append(f" {desc}\n")
+            
+            # Optional keys note
+            content.append("\n⚙️  Optional Keys (disabled by default)\n", style="bold cyan")
+            content.append("─" * 40 + "\n", style="dim")
+            content.append("  Ctrl keys: ", style="dim")
+            content.append("PRAISONAI_TUI_CTRL_KEYS=1\n", style="yellow")
+            content.append("  Fn keys:   ", style="dim")
+            content.append("PRAISONAI_TUI_FN_KEYS=1\n", style="yellow")
+            
+            return content
+        
+        def on_button_pressed(self, event: Button.Pressed) -> None:
+            """Handle button press."""
+            if event.button.id == "close-button":
+                self.dismiss()
+        
+        def action_dismiss(self) -> None:
+            """Dismiss the help screen."""
+            self.dismiss()
+
+else:
+    class HelpScreen:
+        """Placeholder when Textual is not available."""
+        def __init__(self, *args, **kwargs):
+            raise ImportError(
+                "Textual is required for TUI. Install with: pip install praisonai[tui]"
+            )
@@ -25,6 +25,9 @@ if TEXTUAL_AVAILABLE:
    from ..widgets.status import StatusWidget, StatusInfo
    from ..widgets.queue_panel import QueuePanelWidget
    from ..widgets.tool_panel import ToolPanelWidget
+    from ..widgets.command_popup import CommandPopupWidget, get_default_commands
+    from ..widgets.file_popup import FilePopupWidget
+    from ..config import TUIConfig, get_command

    class MainScreen(Screen):
        """
@@ -39,16 +42,18 @@ if TEXTUAL_AVAILABLE:
        - Footer with keybindings
        """
        
+        # Safe default bindings - no Ctrl or Fn keys by default
+        # These work in VS Code, Windsurf, tmux, and plain terminals
        BINDINGS = [
-            Binding("ctrl+q", "quit", "Quit", show=True),
-            Binding("ctrl+c", "cancel", "Cancel", show=True),
-            Binding("ctrl+l", "clear", "Clear", show=True),
-            Binding("f1", "help", "Help", show=True),
-            Binding("f2", "toggle_queue", "Queue", show=True),
-            Binding("f3", "settings", "Settings", show=True),
-            Binding("f5", "clear_chat", "Clear Chat", show=True),
+            # Single-character shortcuts (safe defaults)
+            Binding("q", "quit", "Quit", show=True, priority=True),
+            Binding("question_mark", "help", "Help", show=True),
+            Binding("backslash", "show_commands", "\\cmds", show=True),
+            Binding("colon", "command_mode", ":cmd", show=False),
+            Binding("slash", "search", "/search", show=False),
+            # Navigation
            Binding("tab", "focus_next", "Next", show=False),
-            Binding("escape", "cancel_input", "Cancel", show=False),
+            Binding("escape", "cancel_or_dismiss", "Cancel", show=True),
        ]
        
        DEFAULT_CSS = """
@@ -57,6 +62,7 @@ if TEXTUAL_AVAILABLE:
            grid-size: 2;
            grid-columns: 3fr 1fr;
            grid-rows: 1 1fr auto 1;
+            layers: base popup;
        }
        
        MainScreen #status-bar {
@@ -84,6 +90,18 @@ if TEXTUAL_AVAILABLE:
        MainScreen .sidebar-panel {
            height: 1fr;
        }
+        
+        MainScreen #command-popup {
+            layer: popup;
+            dock: bottom;
+            margin: 0 2 2 2;
+        }
+        
+        MainScreen #file-popup {
+            layer: popup;
+            dock: bottom;
+            margin: 0 2 2 2;
+        }
        """
        
        class MessageSubmitted(Message):
@@ -107,6 +125,7 @@ if TEXTUAL_AVAILABLE:
            self,
            show_queue: bool = True,
            show_tools: bool = True,
+            config: Optional[TUIConfig] = None,
            name: Optional[str] = None,
            id: Optional[str] = None,
            classes: Optional[str] = None,
@@ -114,6 +133,15 @@ if TEXTUAL_AVAILABLE:
            super().__init__(name=name, id=id, classes=classes)
            self._show_queue = show_queue
            self._show_tools = show_tools
+            self._config = config or TUIConfig.from_env()
+            self._command_mode = False
+            self._command_buffer = ""
+            self._command_popup_visible = False
+            self._file_popup_visible = False
+            self._workspace = None
+            
+            # Dynamically add Ctrl/Fn bindings if enabled
+            self._setup_optional_bindings()
        
        def compose(self):
            """Compose the screen."""
@@ -131,6 +159,21 @@ if TEXTUAL_AVAILABLE:
            
            yield Footer()
        
+        def _setup_optional_bindings(self) -> None:
+            """Setup optional Ctrl/Fn key bindings based on config."""
+            if self._config.enable_ctrl_keys:
+                # Add Ctrl key bindings (opt-in)
+                self._bindings.bind("ctrl+q", "quit", "^Q Quit", show=True)
+                self._bindings.bind("ctrl+c", "cancel", "^C Cancel", show=True)
+                self._bindings.bind("ctrl+l", "clear_chat", "^L Clear", show=True)
+            
+            if self._config.enable_fn_keys:
+                # Add Function key bindings (opt-in)
+                self._bindings.bind("f1", "help", "F1 Help", show=True)
+                self._bindings.bind("f2", "toggle_queue", "F2 Queue", show=True)
+                self._bindings.bind("f3", "settings", "F3 Settings", show=True)
+                self._bindings.bind("f5", "clear_chat", "F5 Clear", show=True)
+        
        def on_mount(self) -> None:
            """Handle mount."""
            # Focus composer
@@ -140,14 +183,52 @@ if TEXTUAL_AVAILABLE:
        def on_composer_widget_submitted(self, event: ComposerWidget.Submitted) -> None:
            """Handle message submission."""
            if event.is_command:
-                # Parse command
-                parts = event.content[1:].split(maxsplit=1)
+                # Parse command (supports both / and : prefix)
+                content = event.content
+                if content.startswith("/") or content.startswith(":"):
+                    content = content[1:]  # Remove prefix
+                
+                parts = content.split(maxsplit=1)
                command = parts[0] if parts else ""
                args = parts[1] if len(parts) > 1 else ""
+                
+                # Handle local screen commands first
+                if self._handle_local_command(command, args):
+                    return
+                
+                # Forward to app for other commands
                self.post_message(self.CommandExecuted(command, args))
            else:
                self.post_message(self.MessageSubmitted(event.content))
        
+        def _handle_local_command(self, command: str, args: str) -> bool:
+            """Handle commands that can be processed locally without app involvement."""
+            cmd = get_command(command)
+            
+            if cmd == "quit":
+                self.app.exit()
+                return True
+            elif cmd == "clear":
+                self.action_clear_chat()
+                return True
+            elif cmd == "help":
+                self.action_help()
+                return True
+            elif cmd == "tools":
+                self.action_toggle_tools()
+                return True
+            elif cmd == "queue":
+                self.action_toggle_queue()
+                return True
+            elif cmd == "settings":
+                self.action_settings()
+                return True
+            elif cmd == "cancel":
+                self.action_cancel()
+                return True
+            
+            return False  # Not handled locally
+        
        def action_quit(self) -> None:
            """Quit the application."""
            self.app.exit()
@@ -186,6 +267,183 @@ if TEXTUAL_AVAILABLE:
            composer = self.query_one("#composer", ComposerWidget)
            composer.action_cancel()
        
+        def action_command_mode(self) -> None:
+            """Enter command mode - focus composer with : prefix."""
+            composer = self.query_one("#composer", ComposerWidget)
+            composer.focus_input()
+            composer.set_text(":")
+        
+        def action_search(self) -> None:
+            """Enter search mode - focus composer with / prefix."""
+            composer = self.query_one("#composer", ComposerWidget)
+            composer.focus_input()
+            composer.set_text("/")
+        
+        def action_toggle_tools(self) -> None:
+            """Toggle tools panel visibility."""
+            try:
+                tool_panel = self.query_one("#tool-panel", ToolPanelWidget)
+                tool_panel.display = not tool_panel.display
+            except Exception:
+                pass
+        
+        def action_show_commands(self) -> None:
+            """Show command popup (triggered by backslash)."""
+            if self._command_popup_visible:
+                return
+            
+            self._command_popup_visible = True
+            
+            # Create and mount command popup
+            popup = CommandPopupWidget(
+                commands=get_default_commands(),
+                id="command-popup",
+            )
+            self.mount(popup)
+            popup.focus_search()
+        
+        def action_cancel_or_dismiss(self) -> None:
+            """Cancel input or dismiss popup."""
+            if self._command_popup_visible:
+                self._dismiss_command_popup()
+            else:
+                self.action_cancel_input()
+        
+        def _dismiss_command_popup(self) -> None:
+            """Dismiss the command popup."""
+            if not self._command_popup_visible:
+                return
+            
+            try:
+                popup = self.query_one("#command-popup", CommandPopupWidget)
+                popup.remove()
+            except Exception:
+                pass
+            
+            self._command_popup_visible = False
+            
+            # Return focus to composer
+            composer = self.query_one("#composer", ComposerWidget)
+            composer.focus_input()
+        
+        def on_command_popup_widget_command_selected(
+            self, event: CommandPopupWidget.CommandSelected
+        ) -> None:
+            """Handle command selection from popup."""
+            self._dismiss_command_popup()
+            
+            # Execute the selected command
+            if self._handle_local_command(event.command, event.args):
+                return
+            
+            # Forward to app for other commands
+            self.post_message(self.CommandExecuted(event.command, event.args))
+        
+        def on_command_popup_widget_dismissed(
+            self, event: CommandPopupWidget.Dismissed
+        ) -> None:
+            """Handle popup dismissal."""
+            self._dismiss_command_popup()
+        
+        # Auto-trigger event handlers from ComposerWidget
+        
+        def on_composer_widget_backslash_typed(
+            self, event: ComposerWidget.BackslashTyped
+        ) -> None:
+            """Handle backslash typed - auto-show command popup."""
+            # Clear the backslash from composer and show popup
+            composer = self.query_one("#composer", ComposerWidget)
+            composer.set_text("")
+            self.action_show_commands()
+        
+        def on_composer_widget_at_typed(
+            self, event: ComposerWidget.AtTyped
+        ) -> None:
+            """Handle @ typed - DO NOT block typing.
+            
+            MINIMUM VIABLE UX: User can freely type @path/to/file.
+            The popup is optional and must NOT steal focus or block input.
+            For now, we simply allow typing to continue uninterrupted.
+            """
+            # DO NOT show popup that blocks typing
+            # User can manually type: @README.md, @src/main.py, etc.
+            pass
+        
+        def on_composer_widget_slash_typed(
+            self, event: ComposerWidget.SlashTyped
+        ) -> None:
+            """Handle / typed - auto-show command popup with filter."""
+            if not self._command_popup_visible:
+                self._command_popup_visible = True
+                popup = CommandPopupWidget(
+                    commands=get_default_commands(),
+                    id="command-popup",
+                )
+                self.mount(popup)
+                # Set filter to current query
+                popup.filter_text = event.query
+        
+        def _show_file_popup(self, query: str, at_pos: int) -> None:
+            """Show file autocomplete popup."""
+            if self._file_popup_visible:
+                # Update existing popup
+                try:
+                    popup = self.query_one("#file-popup", FilePopupWidget)
+                    popup.set_query(query, at_pos)
+                except Exception:
+                    pass
+                return
+            
+            self._file_popup_visible = True
+            
+            # Create and mount file popup
+            import os
+            popup = FilePopupWidget(
+                root_dir=self._workspace or os.getcwd(),
+                at_pos=at_pos,
+                id="file-popup",
+            )
+            self.mount(popup)
+            popup.set_query(query, at_pos)
+        
+        def _dismiss_file_popup(self) -> None:
+            """Dismiss the file popup."""
+            if not self._file_popup_visible:
+                return
+            
+            try:
+                popup = self.query_one("#file-popup", FilePopupWidget)
+                popup.remove()
+            except Exception:
+                pass
+            
+            self._file_popup_visible = False
+            
+            # Return focus to composer
+            composer = self.query_one("#composer", ComposerWidget)
+            composer.focus_input()
+        
+        def on_file_popup_widget_file_selected(
+            self, event: FilePopupWidget.FileSelected
+        ) -> None:
+            """Handle file selection from popup."""
+            self._dismiss_file_popup()
+            
+            # Insert the file path into composer
+            composer = self.query_one("#composer", ComposerWidget)
+            current_text = composer.text
+            
+            # Replace from @ position to end with selected path
+            new_text = current_text[:event.at_pos] + "@" + event.path
+            composer.set_text(new_text)
+            composer.focus_input()
+        
+        def on_file_popup_widget_dismissed(
+            self, event: FilePopupWidget.Dismissed
+        ) -> None:
+            """Handle file popup dismissal."""
+            self._dismiss_file_popup()
+        
        # Public methods for updating UI
        
        async def add_user_message(self, content: str) -> None:
@@ -0,0 +1,181 @@
+"""
+Session Store for PraisonAI TUI.
+
+Maintains chat history in memory for session persistence.
+This allows the AI to remember previous messages within a session.
+"""
+
+import logging
+from dataclasses import dataclass, field
+from typing import Dict, List, Optional
+from datetime import datetime
+
+logger = logging.getLogger(__name__)
+
+
+@dataclass
+class ChatMessage:
+    """A single chat message."""
+    role: str  # "user" or "assistant"
+    content: str
+    timestamp: datetime = field(default_factory=datetime.now)
+    run_id: Optional[str] = None
+
+
+@dataclass
+class Session:
+    """A chat session with history."""
+    session_id: str
+    messages: List[ChatMessage] = field(default_factory=list)
+    created_at: datetime = field(default_factory=datetime.now)
+    updated_at: datetime = field(default_factory=datetime.now)
+    
+    def add_message(self, role: str, content: str, run_id: Optional[str] = None) -> None:
+        """Add a message to the session."""
+        self.messages.append(ChatMessage(
+            role=role,
+            content=content,
+            run_id=run_id,
+        ))
+        self.updated_at = datetime.now()
+    
+    def get_history_for_llm(self, max_messages: int = 50) -> List[Dict[str, str]]:
+        """
+        Get chat history in LLM-compatible format.
+        
+        Returns list of {"role": "user/assistant", "content": "..."} dicts.
+        """
+        # Get last N messages
+        recent = self.messages[-max_messages:] if len(self.messages) > max_messages else self.messages
+        return [{"role": msg.role, "content": msg.content} for msg in recent]
+    
+    @property
+    def run_count(self) -> int:
+        """Number of user messages (runs) in this session."""
+        return sum(1 for msg in self.messages if msg.role == "user")
+
+
+class SessionStore:
+    """
+    In-memory session store for TUI.
+    
+    Maintains chat history across agent runs within a session.
+    """
+    
+    def __init__(self, max_sessions: int = 100, max_messages_per_session: int = 1000):
+        """
+        Initialize session store.
+        
+        Args:
+            max_sessions: Maximum number of sessions to keep in memory.
+            max_messages_per_session: Maximum messages per session.
+        """
+        self._sessions: Dict[str, Session] = {}
+        self._max_sessions = max_sessions
+        self._max_messages = max_messages_per_session
+    
+    def get_or_create_session(self, session_id: str) -> Session:
+        """Get existing session or create new one."""
+        if session_id not in self._sessions:
+            self._sessions[session_id] = Session(session_id=session_id)
+            self._cleanup_old_sessions()
+            logger.debug(f"Created new session: {session_id}")
+        return self._sessions[session_id]
+    
+    def add_user_message(self, session_id: str, content: str, run_id: Optional[str] = None) -> None:
+        """Add a user message to the session."""
+        session = self.get_or_create_session(session_id)
+        session.add_message("user", content, run_id)
+        self._trim_session(session)
+        logger.debug(f"Added user message to session {session_id}: {content[:50]}...")
+    
+    def add_assistant_message(self, session_id: str, content: str, run_id: Optional[str] = None) -> None:
+        """Add an assistant message to the session."""
+        session = self.get_or_create_session(session_id)
+        session.add_message("assistant", content, run_id)
+        self._trim_session(session)
+        logger.debug(f"Added assistant message to session {session_id}: {content[:50]}...")
+    
+    def get_chat_history(self, session_id: str, max_messages: int = 50) -> List[Dict[str, str]]:
+        """
+        Get chat history for a session in LLM-compatible format.
+        
+        Args:
+            session_id: The session ID.
+            max_messages: Maximum number of messages to return.
+            
+        Returns:
+            List of {"role": "user/assistant", "content": "..."} dicts.
+        """
+        if session_id not in self._sessions:
+            return []
+        return self._sessions[session_id].get_history_for_llm(max_messages)
+    
+    def get_session(self, session_id: str) -> Optional[Session]:
+        """Get a session by ID."""
+        return self._sessions.get(session_id)
+    
+    def list_sessions(self, limit: int = 50) -> List[Dict]:
+        """List all sessions with metadata."""
+        sessions = sorted(
+            self._sessions.values(),
+            key=lambda s: s.updated_at,
+            reverse=True
+        )[:limit]
+        
+        return [
+            {
+                "session_id": s.session_id,
+                "created_at": s.created_at.strftime("%Y-%m-%d %H:%M"),
+                "updated_at": s.updated_at.strftime("%Y-%m-%d %H:%M"),
+                "run_count": s.run_count,
+            }
+            for s in sessions
+        ]
+    
+    def delete_session(self, session_id: str) -> bool:
+        """Delete a session."""
+        if session_id in self._sessions:
+            del self._sessions[session_id]
+            logger.debug(f"Deleted session: {session_id}")
+            return True
+        return False
+    
+    def clear_session(self, session_id: str) -> None:
+        """Clear all messages from a session."""
+        if session_id in self._sessions:
+            self._sessions[session_id].messages.clear()
+            self._sessions[session_id].updated_at = datetime.now()
+            logger.debug(f"Cleared session: {session_id}")
+    
+    def _trim_session(self, session: Session) -> None:
+        """Trim session to max messages."""
+        if len(session.messages) > self._max_messages:
+            # Keep the most recent messages
+            session.messages = session.messages[-self._max_messages:]
+    
+    def _cleanup_old_sessions(self) -> None:
+        """Remove oldest sessions if over limit."""
+        if len(self._sessions) > self._max_sessions:
+            # Sort by updated_at and remove oldest
+            sorted_sessions = sorted(
+                self._sessions.items(),
+                key=lambda x: x[1].updated_at
+            )
+            # Remove oldest sessions
+            to_remove = len(self._sessions) - self._max_sessions
+            for session_id, _ in sorted_sessions[:to_remove]:
+                del self._sessions[session_id]
+                logger.debug(f"Removed old session: {session_id}")
+
+
+# Global session store instance for TUI
+_session_store: Optional[SessionStore] = None
+
+
+def get_session_store() -> SessionStore:
+    """Get the global session store instance."""
+    global _session_store
+    if _session_store is None:
+        _session_store = SessionStore()
+    return _session_store
@@ -10,9 +10,8 @@ import time

 try:
    from textual.widget import Widget
-    from textual.widgets import Static, RichLog
-    from textual.containers import ScrollableContainer, Vertical
-    from textual.reactive import reactive
+    from textual.widgets import Static
+    from textual.containers import VerticalScroll
    from textual.message import Message
    from rich.markdown import Markdown
    from rich.panel import Panel
@@ -49,23 +48,21 @@ class ChatMessage:


 if TEXTUAL_AVAILABLE:
-    class ChatWidget(ScrollableContainer):
+    class ChatWidget(VerticalScroll):
        """
-        Widget for displaying chat history.
+        Widget for displaying chat history with scrollbar.
        
-        Features:
-        - Streaming message updates
-        - Markdown rendering
-        - Auto-scroll to bottom
-        - Message history management
+        Uses VerticalScroll for proper scrollbar support.
+        Messages are mounted directly to this container.
        """
        
        DEFAULT_CSS = """
        ChatWidget {
-            height: 100%;
+            height: 1fr;
            border: solid $primary;
            background: $surface;
            padding: 0 1;
+            overflow-y: auto;
        }
        
        ChatWidget .message-user {
@@ -118,29 +115,30 @@ if TEXTUAL_AVAILABLE:
            self._streaming_widgets: dict = {}
        
        def compose(self):
-            """Compose the widget."""
-            yield Vertical(id="chat-messages")
+            """Compose the widget - no inner container needed."""
+            # Messages are mounted directly to this VerticalScroll
+            # Must yield from empty iterable (not return None)
+            yield from ()
        
        async def add_message(self, message: ChatMessage) -> None:
-            """Add a message to the chat."""
+            """Add a message to the chat.
+            
+            NEW BEHAVIOR: Messages render NEWEST at TOP.
+            This guarantees visibility without relying on scroll.
+            """
            self._messages.append(message)
            
            # Trim old messages if needed
            if len(self._messages) > self._max_messages:
                self._messages = self._messages[-self._max_messages:]
            
-            # Create message widget
+            # Render message and scroll to show it
            await self._render_message(message)
            
-            # Auto-scroll to bottom
-            self.scroll_end(animate=False)
-            
            self.post_message(self.MessageAdded(message))
        
        async def _render_message(self, message: ChatMessage) -> None:
-            """Render a single message."""
-            container = self.query_one("#chat-messages", Vertical)
-            
+            """Render a message and scroll to show it."""
            # Create role label
            role_style = {
                "user": "bold cyan",
@@ -153,7 +151,7 @@ if TEXTUAL_AVAILABLE:
            
            # Create content
            try:
-                content = Markdown(message.content)
+                content = Markdown(message.content) if message.content else Text("")
            except Exception:
                content = Text(message.content)
            
@@ -170,7 +168,11 @@ if TEXTUAL_AVAILABLE:
                classes=css_class,
            )
            
-            await container.mount(panel)
+            # Mount directly to this VerticalScroll container
+            await self.mount(panel)
+            
+            # Scroll to end to show new message
+            self.scroll_end(animate=False)
            
            if message.is_streaming:
                self._streaming_widgets[message.run_id] = widget_id
@@ -199,7 +201,7 @@ if TEXTUAL_AVAILABLE:
                
                widget.update(Panel(rendered, title="Assistant", border_style="bold green"))
                
-                # Auto-scroll
+                # Scroll to end to keep streaming content visible
                self.scroll_end(animate=False)
                
            except Exception:
@@ -239,8 +241,8 @@ if TEXTUAL_AVAILABLE:
            self._messages.clear()
            self._streaming_widgets.clear()
            
-            container = self.query_one("#chat-messages", Vertical)
-            await container.remove_children()
+            # Remove all children from this container
+            await self.remove_children()
        
        @property
        def messages(self) -> List[ChatMessage]:
@@ -0,0 +1,263 @@
+"""
+Command Popup Widget for PraisonAI TUI.
+
+Shows a searchable list of available commands when user types backslash.
+Inspired by Claude Code's command discovery UX.
+"""
+
+from typing import List, Optional, Callable
+from dataclasses import dataclass
+
+try:
+    from textual.widget import Widget
+    from textual.widgets import Static, Input, OptionList
+    from textual.widgets.option_list import Option
+    from textual.containers import Vertical, Container
+    from textual.reactive import reactive
+    from textual.message import Message
+    from textual import events
+    from rich.text import Text
+    TEXTUAL_AVAILABLE = True
+except ImportError:
+    TEXTUAL_AVAILABLE = False
+    Widget = object
+    Message = object
+
+
+@dataclass
+class CommandInfo:
+    """Information about a command for display."""
+    name: str
+    description: str
+    aliases: List[str]
+    category: str = "general"
+    
+    @property
+    def display_name(self) -> str:
+        """Get display name with aliases."""
+        if self.aliases:
+            return f"{self.name} ({', '.join(self.aliases)})"
+        return self.name
+
+
+if TEXTUAL_AVAILABLE:
+    class CommandPopupWidget(Container):
+        """
+        Popup widget for command discovery and selection.
+        
+        Features:
+        - Shows all available commands
+        - Searchable/filterable
+        - Keyboard navigation
+        - Category grouping
+        """
+        
+        DEFAULT_CSS = """
+        CommandPopupWidget {
+            layer: popup;
+            width: 60;
+            height: auto;
+            max-height: 20;
+            background: $surface;
+            border: solid $primary;
+            padding: 1;
+            margin: 0 0 0 2;
+        }
+        
+        CommandPopupWidget #popup-title {
+            height: 1;
+            background: $primary;
+            color: $text;
+            text-align: center;
+            margin-bottom: 1;
+        }
+        
+        CommandPopupWidget #popup-search {
+            height: 3;
+            margin-bottom: 1;
+        }
+        
+        CommandPopupWidget #popup-list {
+            height: auto;
+            max-height: 12;
+            background: $surface-darken-1;
+        }
+        
+        CommandPopupWidget .command-item {
+            padding: 0 1;
+        }
+        
+        CommandPopupWidget .command-item:hover {
+            background: $primary-darken-1;
+        }
+        
+        CommandPopupWidget #popup-hint {
+            height: 1;
+            color: $text-muted;
+            text-align: center;
+            margin-top: 1;
+        }
+        """
+        
+        class CommandSelected(Message):
+            """Event when a command is selected."""
+            def __init__(self, command: str, args: str = ""):
+                self.command = command
+                self.args = args
+                super().__init__()
+        
+        class Dismissed(Message):
+            """Event when popup is dismissed."""
+            pass
+        
+        # Reactive properties
+        filter_text: reactive[str] = reactive("")
+        
+        def __init__(
+            self,
+            commands: Optional[List[CommandInfo]] = None,
+            name: Optional[str] = None,
+            id: Optional[str] = None,
+            classes: Optional[str] = None,
+        ):
+            super().__init__(name=name, id=id, classes=classes)
+            self._commands = commands or []
+            self._filtered_commands: List[CommandInfo] = []
+        
+        def compose(self):
+            """Compose the widget."""
+            yield Static("Commands", id="popup-title")
+            yield Input(placeholder="Type to filter...", id="popup-search")
+            yield OptionList(id="popup-list")
+            yield Static("↑↓ Navigate • Enter Select • Esc Cancel", id="popup-hint")
+        
+        def on_mount(self) -> None:
+            """Handle mount."""
+            self._update_list()
+            # Focus the search input
+            search = self.query_one("#popup-search", Input)
+            search.focus()
+        
+        def set_commands(self, commands: List[CommandInfo]) -> None:
+            """Set the available commands."""
+            self._commands = commands
+            self._update_list()
+        
+        def watch_filter_text(self, value: str) -> None:
+            """React to filter text changes."""
+            self._update_list()
+        
+        def _update_list(self) -> None:
+            """Update the command list based on filter."""
+            filter_lower = self.filter_text.lower()
+            
+            if filter_lower:
+                self._filtered_commands = [
+                    cmd for cmd in self._commands
+                    if (filter_lower in cmd.name.lower() or
+                        filter_lower in cmd.description.lower() or
+                        any(filter_lower in alias.lower() for alias in cmd.aliases))
+                ]
+            else:
+                self._filtered_commands = self._commands.copy()
+            
+            # Update the option list
+            try:
+                option_list = self.query_one("#popup-list", OptionList)
+                option_list.clear_options()
+                
+                for cmd in self._filtered_commands:
+                    # Create rich text for the option
+                    text = Text()
+                    text.append(f"/{cmd.name}", style="bold cyan")
+                    if cmd.aliases:
+                        text.append(f" ({', '.join(cmd.aliases)})", style="dim")
+                    text.append(f"\n  {cmd.description}", style="")
+                    
+                    option_list.add_option(Option(text, id=cmd.name))
+            except Exception:
+                pass
+        
+        def on_input_changed(self, event: Input.Changed) -> None:
+            """Handle search input changes."""
+            if event.input.id == "popup-search":
+                self.filter_text = event.value
+        
+        def on_input_submitted(self, event: Input.Submitted) -> None:
+            """Handle Enter in search input."""
+            if event.input.id == "popup-search":
+                # Select first matching command
+                if self._filtered_commands:
+                    self._select_command(self._filtered_commands[0].name)
+        
+        def on_option_list_option_selected(self, event: OptionList.OptionSelected) -> None:
+            """Handle command selection from list."""
+            if event.option.id:
+                self._select_command(str(event.option.id))
+        
+        def on_key(self, event: events.Key) -> None:
+            """Handle key events."""
+            if event.key == "escape":
+                self.post_message(self.Dismissed())
+                event.stop()
+            elif event.key == "down":
+                # Move focus to list if in search
+                try:
+                    option_list = self.query_one("#popup-list", OptionList)
+                    option_list.focus()
+                except Exception:
+                    pass
+            elif event.key == "up":
+                # Move focus to search if at top of list
+                try:
+                    search = self.query_one("#popup-search", Input)
+                    search.focus()
+                except Exception:
+                    pass
+        
+        def _select_command(self, command_name: str) -> None:
+            """Select a command and emit event."""
+            self.post_message(self.CommandSelected(command_name))
+        
+        def focus_search(self) -> None:
+            """Focus the search input."""
+            try:
+                search = self.query_one("#popup-search", Input)
+                search.focus()
+            except Exception:
+                pass
+
+
+else:
+    class CommandPopupWidget:
+        """Placeholder when Textual is not available."""
+        def __init__(self, *args, **kwargs):
+            raise ImportError(
+                "Textual is required for TUI. Install with: pip install praisonai[tui]"
+            )
+
+
+# Default commands list (matches slash_commands.py registry)
+DEFAULT_COMMANDS = [
+    CommandInfo("help", "Show help for commands", ["h", "?"]),
+    CommandInfo("clear", "Clear conversation history", ["reset"]),
+    CommandInfo("model", "Show or change the current model", ["m"]),
+    CommandInfo("cost", "Show session cost and token usage", ["usage", "stats"]),
+    CommandInfo("tokens", "Show token usage breakdown", []),
+    CommandInfo("queue", "Show queue status", ["q"]),
+    CommandInfo("cancel", "Cancel current operation", ["c"]),
+    CommandInfo("settings", "Show current settings", ["set"]),
+    CommandInfo("sessions", "Browse saved sessions", ["sess"]),
+    CommandInfo("tools", "Toggle tools panel", ["t"]),
+    CommandInfo("exit", "Exit the TUI", ["quit", "q"]),
+    CommandInfo("plan", "Create an execution plan", []),
+    CommandInfo("undo", "Undo the last change", []),
+    CommandInfo("diff", "Show git diff of changes", []),
+    CommandInfo("commit", "Commit changes with AI message", []),
+    CommandInfo("map", "Show repository map", ["repo"]),
+]
+
+
+def get_default_commands() -> List[CommandInfo]:
+    """Get the default command list."""
+    return DEFAULT_COMMANDS.copy()
@@ -6,6 +6,7 @@ Input area for composing messages with slash command support.

 from typing import Callable, List, Optional
 import asyncio
+import os

 try:
    from textual.widget import Widget
@@ -16,6 +17,31 @@ try:
    from textual import events
    from rich.text import Text
    TEXTUAL_AVAILABLE = True
+    
+    class SubmitTextArea(TextArea):
+        """Custom TextArea that submits on Enter and inserts newline on Shift+Enter."""
+        
+        class SubmitRequested(Message):
+            """Message sent when Enter is pressed."""
+            pass
+        
+        async def _on_key(self, event: events.Key) -> None:
+            """Handle key events before TextArea processes them."""
+            if event.key == "enter":
+                # Don't let TextArea handle Enter - we want to submit
+                event.prevent_default()
+                event.stop()
+                self.post_message(self.SubmitRequested())
+                return
+            elif event.key == "shift+enter":
+                # Insert newline on Shift+Enter
+                event.prevent_default()
+                event.stop()
+                self.insert("\n")
+                return
+            # Let TextArea handle other keys
+            await super()._on_key(event)
+            
 except ImportError:
    TEXTUAL_AVAILABLE = False
    Widget = object
@@ -36,7 +62,7 @@ if TEXTUAL_AVAILABLE:
        """
        
        BINDINGS = [
-            Binding("ctrl+enter", "submit", "Send", show=True),
+            Binding("enter", "submit", "Send", show=True),
            Binding("escape", "cancel", "Cancel", show=False),
            Binding("up", "history_prev", "Previous", show=False),
            Binding("down", "history_next", "Next", show=False),
@@ -77,9 +103,26 @@ if TEXTUAL_AVAILABLE:
                self.args = args
                super().__init__()
        
+        class BackslashTyped(Message):
+            """Event when backslash is typed - triggers command popup."""
+            pass
+        
+        class AtTyped(Message):
+            """Event when @ is typed - triggers file autocomplete."""
+            def __init__(self, query: str, cursor_pos: int):
+                self.query = query
+                self.cursor_pos = cursor_pos
+                super().__init__()
+        
+        class SlashTyped(Message):
+            """Event when / is typed - triggers command discovery."""
+            def __init__(self, query: str):
+                self.query = query
+                super().__init__()
+        
        def __init__(
            self,
-            placeholder: str = "Type your message... (Ctrl+Enter to send, /help for commands)",
+            placeholder: str = "Type your message... (Enter to send, Shift+Enter for newline, /help for commands)",
            multiline: bool = True,
            name: Optional[str] = None,
            id: Optional[str] = None,
@@ -95,7 +138,7 @@ if TEXTUAL_AVAILABLE:
        
        def compose(self):
            """Compose the widget."""
-            yield TextArea(
+            yield SubmitTextArea(
                "",
                id="composer-input",
                language=None,
@@ -103,36 +146,53 @@ if TEXTUAL_AVAILABLE:
        
        def on_mount(self) -> None:
            """Handle mount event."""
-            text_area = self.query_one("#composer-input", TextArea)
+            text_area = self.query_one("#composer-input", SubmitTextArea)
            text_area.focus()
        
+        def on_submit_text_area_submit_requested(self, event: SubmitTextArea.SubmitRequested) -> None:
+            """Handle submit request from TextArea."""
+            self.action_submit()
+        
        async def on_text_area_changed(self, event: TextArea.Changed) -> None:
            """Handle text changes."""
            content = event.text_area.text
            
-            # Detect slash commands
+            # Detect backslash - auto-trigger command popup
+            if content == "\\":
+                self.post_message(self.BackslashTyped())
+                return
+            
+            # Detect @ for file autocomplete
+            if "@" in content:
+                # Find the @ and extract query
+                at_pos = content.rfind("@")
+                if at_pos >= 0:
+                    query = content[at_pos + 1:]
+                    # Only trigger if no space after @
+                    if " " not in query or query.endswith(" ") is False:
+                        self.post_message(self.AtTyped(query, at_pos))
+            
+            # Detect slash commands - auto-trigger command discovery
            if content.startswith("/"):
                parts = content[1:].split(maxsplit=1)
                if parts:
                    command = parts[0]
                    args = parts[1] if len(parts) > 1 else ""
                    self.post_message(self.CommandDetected(command, args))
+                # Also trigger slash popup for discovery
+                self.post_message(self.SlashTyped(content[1:]))
        
        def on_key(self, event: events.Key) -> None:
-            """Handle key events."""
-            # Submit on Enter (single line mode) or Ctrl+Enter
-            if event.key == "enter" and not self._multiline:
-                event.prevent_default()
-                self.action_submit()
-            elif event.key == "ctrl+enter":
-                event.prevent_default()
-                self.action_submit()
+            """Handle key events at widget level (backup handler)."""
+            # These are handled by SubmitTextArea now, but keep as backup
+            pass
        
        def action_submit(self) -> None:
-            """Submit the current input."""
-            if self._is_processing:
-                return
+            """Submit the current input.
            
+            Note: We allow submitting even while processing to enable
+            queueing multiple messages.
+            """
            text_area = self.query_one("#composer-input", TextArea)
            content = text_area.text.strip()
            
@@ -144,8 +204,8 @@ if TEXTUAL_AVAILABLE:
                self._history.append(content)
            self._history_index = -1
            
-            # Check if it's a command
-            is_command = content.startswith("/")
+            # Check if it's a command (/ or : prefix)
+            is_command = content.startswith("/") or content.startswith(":")
            
            # Clear input
            text_area.clear()
@@ -192,10 +252,12 @@ if TEXTUAL_AVAILABLE:
                text_area.insert(self._current_input)
        
        def set_processing(self, processing: bool) -> None:
-            """Set processing state."""
+            """Set processing state.
+            
+            Note: We no longer disable the text area during processing
+            to allow users to queue additional messages while waiting.
+            """
            self._is_processing = processing
-            text_area = self.query_one("#composer-input", TextArea)
-            text_area.disabled = processing
        
        @property
        def text(self) -> str:
@@ -214,6 +276,12 @@ if TEXTUAL_AVAILABLE:
            """Focus the input area."""
            text_area = self.query_one("#composer-input", TextArea)
            text_area.focus()
+        
+        def set_text(self, value: str) -> None:
+            """Set the text content and move cursor to end."""
+            text_area = self.query_one("#composer-input", SubmitTextArea)
+            text_area.clear()
+            text_area.insert(value)

 else:
    class ComposerWidget:
@@ -0,0 +1,207 @@
+"""
+File Autocomplete Popup Widget for PraisonAI TUI.
+
+Shows file/directory suggestions when user types @.
+Reuses FileSearchService from at_mentions.py.
+"""
+
+from typing import List, Optional
+from dataclasses import dataclass
+import os
+
+try:
+    from textual.widget import Widget
+    from textual.widgets import Static, Input, OptionList
+    from textual.widgets.option_list import Option
+    from textual.containers import Container
+    from textual.reactive import reactive
+    from textual.message import Message
+    from textual import events
+    from rich.text import Text
+    TEXTUAL_AVAILABLE = True
+except ImportError:
+    TEXTUAL_AVAILABLE = False
+    Widget = object
+    Message = object
+
+
+@dataclass
+class FileInfo:
+    """Information about a file for display."""
+    path: str
+    file_type: str  # "file" or "directory"
+    score: int = 100
+    
+    @property
+    def icon(self) -> str:
+        """Get icon for file type."""
+        return "📁 " if self.file_type == "directory" else "📄 "
+    
+    @property
+    def display(self) -> str:
+        """Get display string."""
+        return f"{self.icon}{self.path}"
+
+
+if TEXTUAL_AVAILABLE:
+    # Import FileSearchService from at_mentions
+    try:
+        from ..at_mentions import FileSearchService, detect_at_mention
+        FILE_SEARCH_AVAILABLE = True
+    except ImportError:
+        FILE_SEARCH_AVAILABLE = False
+        FileSearchService = None
+
+    class FilePopupWidget(Container):
+        """
+        Popup widget for file autocomplete.
+        
+        Features:
+        - Shows files/directories matching query
+        - Fuzzy search
+        - Keyboard navigation
+        - Supports ~/ and absolute paths
+        """
+        
+        DEFAULT_CSS = """
+        FilePopupWidget {
+            layer: popup;
+            width: 60;
+            height: auto;
+            max-height: 15;
+            background: $surface;
+            border: solid $accent;
+            padding: 1;
+            margin: 0 0 0 2;
+        }
+        
+        FilePopupWidget #file-popup-title {
+            height: 1;
+            background: $accent;
+            color: $text;
+            text-align: center;
+            margin-bottom: 1;
+        }
+        
+        FilePopupWidget #file-popup-list {
+            height: auto;
+            max-height: 10;
+            background: $surface-darken-1;
+        }
+        
+        FilePopupWidget #file-popup-hint {
+            height: 1;
+            color: $text-muted;
+            text-align: center;
+            margin-top: 1;
+        }
+        """
+        
+        class FileSelected(Message):
+            """Event when a file is selected."""
+            def __init__(self, path: str, at_pos: int):
+                self.path = path
+                self.at_pos = at_pos
+                super().__init__()
+        
+        class Dismissed(Message):
+            """Event when popup is dismissed."""
+            pass
+        
+        # Reactive properties
+        filter_text: reactive[str] = reactive("")
+        
+        def __init__(
+            self,
+            root_dir: Optional[str] = None,
+            at_pos: int = 0,
+            name: Optional[str] = None,
+            id: Optional[str] = None,
+            classes: Optional[str] = None,
+        ):
+            super().__init__(name=name, id=id, classes=classes)
+            self._root_dir = root_dir or os.getcwd()
+            self._at_pos = at_pos
+            self._files: List[FileInfo] = []
+            self._file_service = None
+            if FILE_SEARCH_AVAILABLE and FileSearchService:
+                self._file_service = FileSearchService(self._root_dir)
+        
+        def compose(self):
+            """Compose the widget."""
+            yield Static("📂 Files", id="file-popup-title")
+            yield OptionList(id="file-popup-list")
+            yield Static("↑↓ Navigate • Enter Select • Esc Cancel", id="file-popup-hint")
+        
+        def on_mount(self) -> None:
+            """Handle mount."""
+            self._update_list()
+        
+        def set_query(self, query: str, at_pos: int) -> None:
+            """Set the search query."""
+            self.filter_text = query
+            self._at_pos = at_pos
+        
+        def watch_filter_text(self, value: str) -> None:
+            """React to filter text changes."""
+            self._update_list()
+        
+        def _update_list(self) -> None:
+            """Update the file list based on filter."""
+            if not self._file_service:
+                return
+            
+            # Search files
+            results = self._file_service.search(self.filter_text, max_results=15)
+            
+            self._files = [
+                FileInfo(
+                    path=r.path,
+                    file_type=r.file_type,
+                    score=r.score
+                )
+                for r in results
+            ]
+            
+            # Update the option list
+            try:
+                option_list = self.query_one("#file-popup-list", OptionList)
+                option_list.clear_options()
+                
+                for file_info in self._files:
+                    text = Text()
+                    text.append(file_info.icon, style="bold")
+                    text.append(file_info.path, style="cyan" if file_info.file_type == "directory" else "")
+                    
+                    option_list.add_option(Option(text, id=file_info.path))
+            except Exception:
+                pass
+        
+        def on_option_list_option_selected(self, event: OptionList.OptionSelected) -> None:
+            """Handle file selection from list."""
+            if event.option.id:
+                self._select_file(str(event.option.id))
+        
+        def on_key(self, event: events.Key) -> None:
+            """Handle key events."""
+            if event.key == "escape":
+                self.post_message(self.Dismissed())
+                event.stop()
+            elif event.key == "enter":
+                # Select first item if available
+                if self._files:
+                    self._select_file(self._files[0].path)
+                event.stop()
+        
+        def _select_file(self, path: str) -> None:
+            """Select a file and emit event."""
+            self.post_message(self.FileSelected(path, self._at_pos))
+
+
+else:
+    class FilePopupWidget:
+        """Placeholder when Textual is not available."""
+        def __init__(self, *args, **kwargs):
+            raise ImportError(
+                "Textual is required for TUI. Install with: pip install praisonai[tui]"
+            )
@@ -93,6 +93,19 @@ if TEXTUAL_AVAILABLE:
                self.run_id = run_id
                super().__init__()
        
+        class DeleteRequested(Message):
+            """Event when delete is requested."""
+            def __init__(self, run_id: str):
+                self.run_id = run_id
+                super().__init__()
+        
+        class EditRequested(Message):
+            """Event when edit is requested."""
+            def __init__(self, run_id: str, new_content: str):
+                self.run_id = run_id
+                self.new_content = new_content
+                super().__init__()
+        
        queued_count: reactive[int] = reactive(0)
        running_count: reactive[int] = reactive(0)
        
@@ -186,6 +199,16 @@ if TEXTUAL_AVAILABLE:
            if self._selected_run_id:
                self.post_message(self.RetryRequested(self._selected_run_id))
        
+        def delete_selected(self) -> None:
+            """Delete the selected run."""
+            if self._selected_run_id:
+                self.post_message(self.DeleteRequested(self._selected_run_id))
+        
+        def edit_selected(self, new_content: str) -> None:
+            """Edit the selected run's input content."""
+            if self._selected_run_id:
+                self.post_message(self.EditRequested(self._selected_run_id, new_content))
+        
        @property
        def selected_run_id(self) -> Optional[str]:
            """Get the selected run ID."""
@@ -4,14 +4,14 @@ Tool Panel Widget for PraisonAI TUI.
 Displays tool execution status and approval dialogs.
 """

-from typing import List, Optional
+from typing import Any, List, Optional
 from dataclasses import dataclass, field
 import time

 try:
    from textual.widget import Widget
    from textual.widgets import Static, ListView, ListItem
-    from textual.containers import Vertical
+    from textual.containers import Vertical, ScrollableContainer
    from textual.reactive import reactive
    from textual.message import Message
    from rich.text import Text
@@ -61,6 +61,17 @@ if TEXTUAL_AVAILABLE:
            padding: 0 1;
        }
        
+        ToolPanelWidget #available-tools-scroll {
+            height: 1fr;
+            min-height: 5;
+            scrollbar-gutter: stable;
+        }
+        
+        ToolPanelWidget #tool-list-scroll {
+            height: auto;
+            max-height: 10;
+        }
+        
        ToolPanelWidget .tool-pending {
            color: $warning;
        }
@@ -82,6 +93,11 @@ if TEXTUAL_AVAILABLE:
            border: solid $warning;
            padding: 1;
        }
+        
+        ToolPanelWidget .tools-count {
+            color: $text-muted;
+            text-style: italic;
+        }
        """
        
        class ApprovalResponse(Message):
@@ -102,15 +118,58 @@ if TEXTUAL_AVAILABLE:
            self._calls: List[ToolCall] = []
            self._max_items = max_items
            self._pending_approval: Optional[ToolCall] = None
+            self._available_tools: List[str] = []
        
        def compose(self):
            """Compose the widget."""
            yield Static("Tools", id="tool-title", classes="tool-header")
-            yield Vertical(id="tool-list")
+            yield ScrollableContainer(Vertical(id="tool-list"), id="tool-list-scroll")
+            yield Static("", id="tools-count", classes="tools-count")
+            yield ScrollableContainer(Vertical(id="available-tools"), id="available-tools-scroll")
        
        def on_mount(self) -> None:
            """Handle mount."""
            self._update_display()
+            self._update_available_tools_display()
+        
+        def set_available_tools(self, tools: List[Any]) -> None:
+            """Set the list of available tools to display."""
+            self._available_tools = []
+            for tool in tools:
+                if hasattr(tool, '__name__'):
+                    self._available_tools.append(tool.__name__)
+                elif hasattr(tool, 'name'):
+                    self._available_tools.append(tool.name)
+                elif callable(tool):
+                    self._available_tools.append(str(tool))
+                else:
+                    self._available_tools.append(str(type(tool).__name__))
+            self._update_available_tools_display()
+        
+        def _update_available_tools_display(self) -> None:
+            """Update the available tools display."""
+            try:
+                container = self.query_one("#available-tools", Vertical)
+                count_widget = self.query_one("#tools-count", Static)
+            except Exception:
+                return
+            
+            container.remove_children()
+            
+            if not self._available_tools:
+                container.mount(Static(Text("No tools loaded", style="dim italic")))
+                count_widget.update("")
+                return
+            
+            # Update count display
+            count_widget.update(f"{len(self._available_tools)} tools available")
+            
+            # Show ALL tools with scrolling support
+            for tool_name in sorted(self._available_tools):
+                text = Text()
+                text.append("• ", style="cyan")
+                text.append(tool_name, style="white")
+                container.mount(Static(text))
        
        def add_call(self, call: ToolCall) -> None:
            """Add a tool call."""
@@ -27,7 +27,7 @@ LEGACY_COMMANDS = {
 TYPER_COMMANDS = {
    'config', 'traces', 'env', 'session', 'completion', 'version',
    'debug', 'lsp', 'diag', 'doctor', 'acp', 'mcp', 'serve', 'schedule', 'run',
-    'tui', 'queue',  # TUI and queue management commands
+    'tui', 'queue', 'profile',  # TUI, queue management, and profiling commands
 }


@@ -329,8 +329,7 @@ class PraisonAI:
                if combined_inputs:
                    combined_prompt = f"{args.command} {' '.join(combined_inputs)}"
                    result = self.handle_direct_prompt(combined_prompt)
-                    if result is not None:
-                        print(result)
+                    # Result already printed by handle_direct_prompt, don't print again
                    return result
                else:
                    self.agent_file = args.command
@@ -357,8 +356,7 @@ class PraisonAI:
                    prompt_parts.append(file_input)
                prompt = ' '.join(prompt_parts)
                result = self.handle_direct_prompt(prompt)
-                if result is not None:
-                    print(result)
+                # Result already printed by handle_direct_prompt, don't print again
                return result
            else:
                # Agent file was explicitly set, ignore direct prompt and use the file
@@ -374,8 +372,7 @@ class PraisonAI:
                    inputs.append(file_input)
                combined_input = ' '.join(inputs)
                result = self.handle_direct_prompt(combined_input)
-                if result is not None:
-                    print(result)
+                # Result already printed by handle_direct_prompt, don't print again
                return result
        # If no command or direct_prompt, preserve agent_file from constructor (don't overwrite)

@@ -737,6 +734,8 @@ class PraisonAI:
        parser.add_argument("--expand-tools", type=str, help="Tools for prompt expander (e.g., 'internet_search' or path to tools.py)")
        parser.add_argument("--tools", "-t", type=str, help="Path to tools.py file for research agent")
        parser.add_argument("--no-tools", action="store_true", help="Disable default built-in tools (for models that don't support tool calling)")
+        parser.add_argument("--no-acp", action="store_true", help="Disable ACP tools (agentic file operations with plan/approve/apply)")
+        parser.add_argument("--no-lsp", action="store_true", help="Disable LSP tools (code intelligence: symbols, definitions, references)")
        parser.add_argument("--save", "-s", action="store_true", help="Save research output to file (output/research/)")
        parser.add_argument("--verbose", "-v", action="store_true", help="Enable verbose output for research")
        parser.add_argument("--web-search", action="store_true", help="Enable native web search (OpenAI, Gemini, Anthropic, xAI, Perplexity)")
@@ -786,6 +785,11 @@ class PraisonAI:
        # Telemetry - usage monitoring
        parser.add_argument("--telemetry", action="store_true", help="Enable usage monitoring and analytics")
        
+        # Profiling - unified execution profiling
+        parser.add_argument("--profile", action="store_true", help="Enable profiling with timing breakdown")
+        parser.add_argument("--profile-deep", action="store_true", dest="profile_deep", help="Enable deep profiling with call graph (higher overhead)")
+        parser.add_argument("--profile-format", type=str, choices=["text", "json"], default="text", dest="profile_format", help="Profile output format")
+        
        # MCP - Model Context Protocol
        parser.add_argument("--mcp", type=str, help="MCP server command (e.g., 'npx -y @modelcontextprotocol/server-filesystem .')")
        parser.add_argument("--mcp-env", type=str, help="MCP environment variables (KEY=value,KEY2=value2)")
@@ -1201,10 +1205,13 @@ class PraisonAI:
                sys.exit(exit_code)
            
            elif args.command == 'profile':
-                # Profile command - performance profiling and benchmarking
-                from .features.profile import handle_profile_command
-                exit_code = handle_profile_command(unknown_args)
-                sys.exit(exit_code)
+                # Profile command - delegate to Typer CLI for new profiler
+                # This routes to commands/profile.py which has query, imports, startup subcommands
+                from .app import app as typer_app
+                import sys as _sys
+                _sys.argv = ['praisonai', 'profile'] + unknown_args
+                typer_app()
+                sys.exit(0)
            
            elif args.command == 'eval':
                # Eval command - evaluate model responses against expected outputs
@@ -3457,6 +3464,10 @@ Provide ONLY the commit message, no explanations."""
        - @rule:name - Include specific rule
        - @url:https://... - Fetch URL content
        """
+        # Check for profiling mode - use unified profiler
+        if hasattr(self, 'args') and getattr(self.args, 'profile', False):
+            return self._handle_profiled_prompt(prompt)
+        
        # Check for inline workflow mode
        if hasattr(self, 'args') and getattr(self.args, 'workflow', None):
            return self._run_inline_workflow(prompt)
@@ -3932,6 +3943,55 @@ Now, {final_instruction.lower()}:"""
            print("pip install praisonaiagents # For PraisonAIAgents\n")  
            sys.exit(1)

+    def _handle_profiled_prompt(self, prompt):
+        """
+        Handle direct prompt with unified profiling enabled.
+        
+        Uses the unified execution module for consistent profiling across
+        CLI direct invocation and profile command.
+        """
+        from .execution import ExecutionRequest, Profiler, ProfilerConfig
+        
+        # Build execution request
+        model = None
+        if hasattr(self, 'args') and self.args.llm:
+            model = self.args.llm
+        
+        request = ExecutionRequest(
+            prompt=prompt,
+            agent_name="DirectAgent",
+            model=model,
+            stream=False,
+        )
+        
+        # Create profiler config from flags
+        deep = getattr(self.args, 'profile_deep', False)
+        output_format = getattr(self.args, 'profile_format', 'text')
+        
+        config = ProfilerConfig(
+            layer=2 if deep else 1,
+            show_callers=deep,
+            show_callees=deep,
+            output_format=output_format,
+        )
+        
+        # Run with profiling
+        print("[bold cyan]🔬 Starting profiled execution...[/bold cyan]")
+        if deep:
+            print("[yellow]⚠️  Deep profiling enabled - higher overhead[/yellow]")
+        
+        profiler = Profiler(config)
+        result, report = profiler.profile_sync(request, invocation_method="cli_direct")
+        
+        # Output profile report
+        if output_format == "json":
+            print(report.to_json())
+        else:
+            print(report.to_text())
+        
+        # Return the actual result for any downstream processing
+        return result.output
+
    def _run_with_status_display(self, agent, prompt):
        """
        Run agent with minimal status display (spinner + tool/handoff updates).
@@ -4883,7 +4943,58 @@ Now, {final_instruction.lower()}:"""
            sys.exit(1)
    
    def _load_interactive_tools(self):
-        """Load tools for interactive mode."""
+        """
+        Load tools for interactive mode using the canonical provider.
+        
+        This method uses the centralized interactive_tools module which provides:
+        - ACP tools (acp_create_file, acp_edit_file, etc.) for safe file operations
+        - LSP tools (lsp_list_symbols, lsp_find_definition, etc.) for code intelligence
+        - Basic tools (read_file, write_file, etc.) for standard operations
+        
+        Tool groups can be disabled via:
+        - CLI flags: --no-acp, --no-lsp
+        - Env vars: PRAISON_TOOLS_DISABLE=acp,lsp
+        """
+        # Determine which groups to disable based on CLI args
+        disable_groups = []
+        if hasattr(self, 'args'):
+            if getattr(self.args, 'no_acp', False):
+                disable_groups.append('acp')
+            if getattr(self.args, 'no_lsp', False):
+                disable_groups.append('lsp')
+        
+        # Get workspace
+        workspace = os.getcwd()
+        
+        try:
+            from .features.interactive_tools import get_interactive_tools, ToolConfig
+            
+            # Create config
+            config = ToolConfig.from_env()
+            config.workspace = workspace
+            
+            # Apply CLI overrides
+            if 'acp' in disable_groups:
+                config.enable_acp = False
+            if 'lsp' in disable_groups:
+                config.enable_lsp = False
+            
+            # Get tools from canonical provider
+            tools_list = get_interactive_tools(
+                config=config,
+                disable=disable_groups if disable_groups else None,
+            )
+            
+            logging.debug(f"Loaded {len(tools_list)} interactive tools (ACP: {config.enable_acp}, LSP: {config.enable_lsp})")
+            return tools_list
+            
+        except ImportError as e:
+            logging.debug(f"Interactive tools provider not available: {e}")
+            # Fallback to basic tools only
+            return self._load_basic_tools_fallback()
+    
+    def _load_basic_tools_fallback(self):
+        """Fallback to load basic tools when interactive_tools module unavailable."""
        tools_list = []
        try:
            from praisonaiagents.tools import (
@@ -57,7 +57,7 @@ class CloudDeployer:
            file.write("FROM python:3.11-slim\n")
            file.write("WORKDIR /app\n")
            file.write("COPY . .\n")
-            file.write("RUN pip install flask praisonai==2.9.2 gunicorn markdown\n")
+            file.write("RUN pip install flask praisonai==3.0.2 gunicorn markdown\n")
            file.write("EXPOSE 8080\n")
            file.write('CMD ["gunicorn", "-b", "0.0.0.0:8080", "api:app"]\n')
            
@@ -1 +1 @@
-__version__ = "2.9.2"
+__version__ = "3.0.2"
@@ -12,7 +12,7 @@ dependencies = [
    "rich>=13.7",
    "markdown>=3.5",
    "pyparsing>=3.0.0",
-    "praisonaiagents>=0.5.3",
+    "praisonaiagents>=0.6.0",
    "python-dotenv>=0.19.0",
    "instructor>=1.3.3",
    "PyYAML>=6.0",
@@ -0,0 +1,287 @@
+"""
+Unit tests for the interactive tools provider.
+
+Tests the canonical source of truth for default tools in interactive modes.
+"""
+
+import os
+import pytest
+from unittest.mock import patch, MagicMock
+
+
+class TestToolGroups:
+    """Tests for tool group definitions."""
+    
+    def test_tool_groups_defined(self):
+        """Test that all expected tool groups are defined."""
+        from praisonai.cli.features.interactive_tools import TOOL_GROUPS
+        
+        assert "acp" in TOOL_GROUPS
+        assert "lsp" in TOOL_GROUPS
+        assert "basic" in TOOL_GROUPS
+        assert "interactive" in TOOL_GROUPS
+    
+    def test_acp_tools_list(self):
+        """Test ACP tool group contains expected tools."""
+        from praisonai.cli.features.interactive_tools import TOOL_GROUPS
+        
+        expected = {
+            "acp_create_file",
+            "acp_edit_file",
+            "acp_delete_file",
+            "acp_execute_command",
+        }
+        assert set(TOOL_GROUPS["acp"]) == expected
+    
+    def test_lsp_tools_list(self):
+        """Test LSP tool group contains expected tools."""
+        from praisonai.cli.features.interactive_tools import TOOL_GROUPS
+        
+        expected = {
+            "lsp_list_symbols",
+            "lsp_find_definition",
+            "lsp_find_references",
+            "lsp_get_diagnostics",
+        }
+        assert set(TOOL_GROUPS["lsp"]) == expected
+    
+    def test_basic_tools_list(self):
+        """Test basic tool group contains expected tools."""
+        from praisonai.cli.features.interactive_tools import TOOL_GROUPS
+        
+        expected = {
+            "read_file",
+            "write_file",
+            "list_files",
+            "execute_command",
+            "internet_search",
+        }
+        assert set(TOOL_GROUPS["basic"]) == expected
+    
+    def test_interactive_group_is_union(self):
+        """Test interactive group is union of all groups."""
+        from praisonai.cli.features.interactive_tools import TOOL_GROUPS
+        
+        expected = set(TOOL_GROUPS["acp"]) | set(TOOL_GROUPS["lsp"]) | set(TOOL_GROUPS["basic"])
+        assert set(TOOL_GROUPS["interactive"]) == expected
+
+
+class TestToolConfig:
+    """Tests for ToolConfig dataclass."""
+    
+    def test_default_config(self):
+        """Test default configuration values."""
+        from praisonai.cli.features.interactive_tools import ToolConfig
+        
+        config = ToolConfig()
+        assert config.enable_acp is True
+        assert config.enable_lsp is True
+        assert config.enable_basic is True
+        assert config.approval_mode == "auto"
+    
+    def test_config_from_env_disable_acp(self):
+        """Test config from env with ACP disabled."""
+        from praisonai.cli.features.interactive_tools import ToolConfig
+        
+        with patch.dict(os.environ, {"PRAISON_TOOLS_DISABLE": "acp"}):
+            config = ToolConfig.from_env()
+            assert config.enable_acp is False
+            assert config.enable_lsp is True
+    
+    def test_config_from_env_disable_lsp(self):
+        """Test config from env with LSP disabled."""
+        from praisonai.cli.features.interactive_tools import ToolConfig
+        
+        with patch.dict(os.environ, {"PRAISON_TOOLS_DISABLE": "lsp"}):
+            config = ToolConfig.from_env()
+            assert config.enable_acp is True
+            assert config.enable_lsp is False
+    
+    def test_config_from_env_disable_multiple(self):
+        """Test config from env with multiple groups disabled."""
+        from praisonai.cli.features.interactive_tools import ToolConfig
+        
+        with patch.dict(os.environ, {"PRAISON_TOOLS_DISABLE": "acp,lsp"}):
+            config = ToolConfig.from_env()
+            assert config.enable_acp is False
+            assert config.enable_lsp is False
+            assert config.enable_basic is True
+    
+    def test_config_from_env_workspace(self):
+        """Test config from env with workspace override."""
+        from praisonai.cli.features.interactive_tools import ToolConfig
+        
+        with patch.dict(os.environ, {"PRAISON_WORKSPACE": "/custom/path"}):
+            config = ToolConfig.from_env()
+            assert config.workspace == "/custom/path"
+
+
+class TestResolveToolGroups:
+    """Tests for resolve_tool_groups function."""
+    
+    def test_resolve_all_enabled(self):
+        """Test resolving with all groups enabled."""
+        from praisonai.cli.features.interactive_tools import resolve_tool_groups, ToolConfig, TOOL_GROUPS
+        
+        config = ToolConfig()
+        tool_names = resolve_tool_groups(config=config)
+        
+        expected = set(TOOL_GROUPS["interactive"])
+        assert tool_names == expected
+    
+    def test_resolve_disable_acp(self):
+        """Test resolving with ACP disabled."""
+        from praisonai.cli.features.interactive_tools import resolve_tool_groups, ToolConfig, TOOL_GROUPS
+        
+        config = ToolConfig(enable_acp=False)
+        tool_names = resolve_tool_groups(disable=["acp"], config=config)
+        
+        # Should not contain ACP tools
+        for acp_tool in TOOL_GROUPS["acp"]:
+            assert acp_tool not in tool_names
+        
+        # Should contain LSP and basic tools
+        for lsp_tool in TOOL_GROUPS["lsp"]:
+            assert lsp_tool in tool_names
+    
+    def test_resolve_disable_lsp(self):
+        """Test resolving with LSP disabled."""
+        from praisonai.cli.features.interactive_tools import resolve_tool_groups, ToolConfig, TOOL_GROUPS
+        
+        config = ToolConfig(enable_lsp=False)
+        tool_names = resolve_tool_groups(disable=["lsp"], config=config)
+        
+        # Should not contain LSP tools
+        for lsp_tool in TOOL_GROUPS["lsp"]:
+            assert lsp_tool not in tool_names
+        
+        # Should contain ACP and basic tools
+        for acp_tool in TOOL_GROUPS["acp"]:
+            assert acp_tool in tool_names
+    
+    def test_resolve_specific_groups(self):
+        """Test resolving specific groups only."""
+        from praisonai.cli.features.interactive_tools import resolve_tool_groups, TOOL_GROUPS
+        
+        tool_names = resolve_tool_groups(groups=["basic"])
+        
+        assert tool_names == set(TOOL_GROUPS["basic"])
+
+
+class TestGetInteractiveTools:
+    """Tests for get_interactive_tools function."""
+    
+    def test_get_tools_returns_list(self):
+        """Test that get_interactive_tools returns a list."""
+        from praisonai.cli.features.interactive_tools import get_interactive_tools, ToolConfig
+        
+        config = ToolConfig(workspace="/tmp")
+        tools = get_interactive_tools(config=config)
+        
+        assert isinstance(tools, list)
+    
+    def test_get_tools_deterministic_order(self):
+        """Test that tool order is deterministic."""
+        from praisonai.cli.features.interactive_tools import get_interactive_tools, ToolConfig
+        
+        config = ToolConfig(workspace="/tmp")
+        tools1 = get_interactive_tools(config=config)
+        tools2 = get_interactive_tools(config=config)
+        
+        names1 = [t.__name__ for t in tools1]
+        names2 = [t.__name__ for t in tools2]
+        
+        assert names1 == names2
+    
+    def test_get_tools_with_disable_acp(self):
+        """Test get_interactive_tools with ACP disabled."""
+        from praisonai.cli.features.interactive_tools import get_interactive_tools, ToolConfig
+        
+        config = ToolConfig(workspace="/tmp", enable_acp=False)
+        tools = get_interactive_tools(config=config, disable=["acp"])
+        
+        tool_names = [t.__name__ for t in tools]
+        acp_tools = [n for n in tool_names if n.startswith("acp_")]
+        
+        assert len(acp_tools) == 0
+    
+    def test_get_tools_with_disable_lsp(self):
+        """Test get_interactive_tools with LSP disabled."""
+        from praisonai.cli.features.interactive_tools import get_interactive_tools, ToolConfig
+        
+        config = ToolConfig(workspace="/tmp", enable_lsp=False)
+        tools = get_interactive_tools(config=config, disable=["lsp"])
+        
+        tool_names = [t.__name__ for t in tools]
+        lsp_tools = [n for n in tool_names if n.startswith("lsp_")]
+        
+        assert len(lsp_tools) == 0
+
+
+class TestMergeToolOverrides:
+    """Tests for merge_tool_overrides function."""
+    
+    def test_merge_add_tools(self):
+        """Test adding tools to default list."""
+        from praisonai.cli.features.interactive_tools import merge_tool_overrides
+        
+        def tool_a():
+            pass
+        
+        def tool_b():
+            pass
+        
+        def tool_c():
+            pass
+        
+        default = [tool_a, tool_b]
+        result = merge_tool_overrides(default, add_tools=[tool_c])
+        
+        names = [t.__name__ for t in result]
+        assert "tool_c" in names
+    
+    def test_merge_remove_tools(self):
+        """Test removing tools from default list."""
+        from praisonai.cli.features.interactive_tools import merge_tool_overrides
+        
+        def tool_a():
+            pass
+        
+        def tool_b():
+            pass
+        
+        default = [tool_a, tool_b]
+        result = merge_tool_overrides(default, remove_names=["tool_a"])
+        
+        names = [t.__name__ for t in result]
+        assert "tool_a" not in names
+        assert "tool_b" in names
+
+
+class TestLazyImports:
+    """Tests for lazy import behavior."""
+    
+    def test_import_time_fast(self):
+        """Test that importing the module is fast (no heavy deps)."""
+        import time
+        
+        start = time.time()
+        # Force reimport
+        import importlib
+        import praisonai.cli.features.interactive_tools as module
+        importlib.reload(module)
+        duration = time.time() - start
+        
+        # Should be under 100ms
+        assert duration < 0.1, f"Import took {duration:.3f}s, expected < 0.1s"
+    
+    def test_tool_groups_available_without_runtime(self):
+        """Test that TOOL_GROUPS is available without runtime initialization."""
+        from praisonai.cli.features.interactive_tools import TOOL_GROUPS
+        
+        # Should be accessible immediately
+        assert len(TOOL_GROUPS) > 0
+
+
+if __name__ == "__main__":
+    pytest.main([__file__, "-v"])
@@ -0,0 +1,366 @@
+"""
+Unit tests for CLI profiler module.
+"""
+
+
+class TestProfilerConfig:
+    """Tests for ProfilerConfig dataclass."""
+    
+    def test_default_config(self):
+        """Test default configuration values."""
+        from praisonai.cli.features.profiler import ProfilerConfig
+        
+        config = ProfilerConfig()
+        assert config.deep is False
+        assert config.limit == 30
+        assert config.sort_by == "cumulative"
+        assert config.show_files is False
+        assert config.show_callers is False
+        assert config.show_callees is False
+        assert config.importtime is False
+        assert config.first_token is False
+        assert config.save_path is None
+        assert config.output_format == "text"
+        assert config.stream is False
+    
+    def test_custom_config(self):
+        """Test custom configuration."""
+        from praisonai.cli.features.profiler import ProfilerConfig
+        
+        config = ProfilerConfig(
+            deep=True,
+            limit=50,
+            sort_by="tottime",
+            show_files=True,
+        )
+        assert config.deep is True
+        assert config.limit == 50
+        assert config.sort_by == "tottime"
+        assert config.show_files is True
+
+
+class TestTimingBreakdown:
+    """Tests for TimingBreakdown dataclass."""
+    
+    def test_default_timing(self):
+        """Test default timing values."""
+        from praisonai.cli.features.profiler.core import TimingBreakdown
+        
+        timing = TimingBreakdown()
+        assert timing.cli_parse_ms == 0.0
+        assert timing.imports_ms == 0.0
+        assert timing.agent_construction_ms == 0.0
+        assert timing.model_init_ms == 0.0
+        assert timing.first_token_ms == 0.0
+        assert timing.total_run_ms == 0.0
+
+
+class TestFunctionStats:
+    """Tests for FunctionStats dataclass."""
+    
+    def test_to_dict(self):
+        """Test conversion to dictionary."""
+        from praisonai.cli.features.profiler.core import FunctionStats
+        
+        stats = FunctionStats(
+            name="test_func",
+            filename="/path/to/file.py",
+            lineno=42,
+            calls=10,
+            tottime=0.5,
+            cumtime=1.0,
+        )
+        
+        result = stats.to_dict()
+        assert result["name"] == "test_func"
+        assert result["filename"] == "/path/to/file.py"
+        assert result["lineno"] == 42
+        assert result["calls"] == 10
+        assert result["tottime_ms"] == 500.0
+        assert result["cumtime_ms"] == 1000.0
+
+
+class TestFileStats:
+    """Tests for FileStats dataclass."""
+    
+    def test_to_dict(self):
+        """Test conversion to dictionary."""
+        from praisonai.cli.features.profiler.core import FileStats, FunctionStats
+        
+        func_stats = FunctionStats(
+            name="func1",
+            filename="/path/to/file.py",
+            lineno=10,
+            calls=5,
+            tottime=0.1,
+            cumtime=0.2,
+        )
+        
+        file_stats = FileStats(
+            filename="/path/to/file.py",
+            total_time=0.5,
+            functions=[func_stats],
+        )
+        
+        result = file_stats.to_dict()
+        assert result["filename"] == "/path/to/file.py"
+        assert result["total_time_ms"] == 500.0
+        assert result["function_count"] == 1
+
+
+class TestProfilerResult:
+    """Tests for ProfilerResult dataclass."""
+    
+    def test_to_dict(self):
+        """Test conversion to dictionary."""
+        from praisonai.cli.features.profiler.core import (
+            ProfilerResult,
+            TimingBreakdown,
+            FunctionStats,
+            FileStats,
+        )
+        
+        timing = TimingBreakdown(
+            cli_parse_ms=1.0,
+            imports_ms=100.0,
+            total_run_ms=500.0,
+        )
+        
+        func_stats = [FunctionStats(
+            name="test",
+            filename="test.py",
+            lineno=1,
+            calls=1,
+            tottime=0.1,
+            cumtime=0.2,
+        )]
+        
+        file_stats = [FileStats(
+            filename="test.py",
+            total_time=0.2,
+        )]
+        
+        result = ProfilerResult(
+            prompt="test prompt",
+            response="test response",
+            timing=timing,
+            function_stats=func_stats,
+            file_stats=file_stats,
+            metadata={"model": "test"},
+        )
+        
+        data = result.to_dict()
+        assert "timestamp" in data
+        assert data["prompt"] == "test prompt"
+        assert data["response_preview"] == "test response"
+        assert data["timing"]["cli_parse_ms"] == 1.0
+        assert data["timing"]["imports_ms"] == 100.0
+        assert len(data["top_functions"]) == 1
+        assert len(data["top_files"]) == 1
+
+
+class TestQueryProfiler:
+    """Tests for QueryProfiler class."""
+    
+    def test_init_default_config(self):
+        """Test initialization with default config."""
+        from praisonai.cli.features.profiler import QueryProfiler, ProfilerConfig
+        
+        profiler = QueryProfiler()
+        assert profiler.config is not None
+        assert isinstance(profiler.config, ProfilerConfig)
+    
+    def test_init_custom_config(self):
+        """Test initialization with custom config."""
+        from praisonai.cli.features.profiler import QueryProfiler, ProfilerConfig
+        
+        config = ProfilerConfig(deep=True, limit=50)
+        profiler = QueryProfiler(config)
+        assert profiler.config.deep is True
+        assert profiler.config.limit == 50
+    
+    def test_collect_metadata(self):
+        """Test metadata collection."""
+        from praisonai.cli.features.profiler import QueryProfiler
+        
+        profiler = QueryProfiler()
+        metadata = profiler._collect_metadata("gpt-4")
+        
+        assert "python_version" in metadata
+        assert "platform" in metadata
+        assert "praisonai_version" in metadata
+        assert metadata["model"] == "gpt-4"
+        assert "timestamp" in metadata
+
+
+class TestFormatProfileReport:
+    """Tests for format_profile_report function."""
+    
+    def test_basic_report(self):
+        """Test basic report formatting."""
+        from praisonai.cli.features.profiler import (
+            format_profile_report,
+            ProfilerConfig,
+        )
+        from praisonai.cli.features.profiler.core import (
+            ProfilerResult,
+            TimingBreakdown,
+            FunctionStats,
+            FileStats,
+        )
+        
+        timing = TimingBreakdown(
+            cli_parse_ms=1.0,
+            imports_ms=100.0,
+            total_run_ms=500.0,
+        )
+        
+        func_stats = [FunctionStats(
+            name="test_func",
+            filename="test.py",
+            lineno=1,
+            calls=10,
+            tottime=0.1,
+            cumtime=0.2,
+        )]
+        
+        file_stats = [FileStats(
+            filename="test.py",
+            total_time=0.2,
+        )]
+        
+        result = ProfilerResult(
+            prompt="test",
+            response="response",
+            timing=timing,
+            function_stats=func_stats,
+            file_stats=file_stats,
+            metadata={"python_version": "3.12", "platform": "test", "praisonai_version": "1.0", "model": "test"},
+        )
+        
+        report = format_profile_report(result)
+        
+        assert "PraisonAI Profile Report" in report
+        assert "System Information" in report
+        assert "Timing Breakdown" in report
+        assert "Per-Function Timing" in report
+        assert "test_func" in report
+    
+    def test_report_with_files(self):
+        """Test report with file grouping."""
+        from praisonai.cli.features.profiler import (
+            format_profile_report,
+            ProfilerConfig,
+        )
+        from praisonai.cli.features.profiler.core import (
+            ProfilerResult,
+            TimingBreakdown,
+            FunctionStats,
+            FileStats,
+        )
+        
+        config = ProfilerConfig(show_files=True)
+        
+        timing = TimingBreakdown()
+        func_stats = []
+        file_stats = [FileStats(filename="test.py", total_time=0.5)]
+        
+        result = ProfilerResult(
+            prompt="test",
+            response="response",
+            timing=timing,
+            function_stats=func_stats,
+            file_stats=file_stats,
+            metadata={},
+        )
+        
+        report = format_profile_report(result, config)
+        assert "Per-File Timing" in report
+
+
+class TestRedactSecrets:
+    """Tests for secret redaction."""
+    
+    def test_redact_openai_key(self):
+        """Test redacting OpenAI API keys."""
+        from praisonai.cli.features.profiler.core import redact_secrets
+        
+        text = "API key: sk-1234567890abcdefghijklmnopqrstuvwxyz"
+        result = redact_secrets(text)
+        assert "sk-1234567890" not in result
+        assert "[REDACTED]" in result
+    
+    def test_redact_anthropic_key(self):
+        """Test redacting Anthropic API keys."""
+        from praisonai.cli.features.profiler.core import redact_secrets
+        
+        text = "Key: sk-ant-api03-abcdefghijklmnopqrstuvwxyz"
+        result = redact_secrets(text)
+        assert "sk-ant-api03" not in result
+        assert "[REDACTED]" in result
+    
+    def test_preserve_normal_text(self):
+        """Test that normal text is preserved."""
+        from praisonai.cli.features.profiler.core import redact_secrets
+        
+        text = "This is normal text without secrets"
+        result = redact_secrets(text)
+        assert result == text
+
+
+class TestCLIImports:
+    """Tests for CLI import functionality."""
+    
+    def test_cli_app_imports(self):
+        """Test that CLI app imports without errors."""
+        # This tests the fix for ModuleNotFoundError
+        from praisonai.cli.app import app
+        assert app is not None
+    
+    def test_output_module_imports(self):
+        """Test that output module imports correctly."""
+        from praisonai.cli.output import OutputController, OutputMode
+        assert OutputController is not None
+        assert OutputMode is not None
+    
+    def test_profiler_module_imports(self):
+        """Test that profiler module imports correctly."""
+        from praisonai.cli.features.profiler import (
+            ProfilerConfig,
+            ProfilerResult,
+            QueryProfiler,
+            run_profiled_query,
+            format_profile_report,
+        )
+        assert ProfilerConfig is not None
+        assert ProfilerResult is not None
+        assert QueryProfiler is not None
+        assert run_profiled_query is not None
+        assert format_profile_report is not None
+
+
+class TestOutputController:
+    """Tests for OutputController."""
+    
+    def test_default_mode(self):
+        """Test default output mode."""
+        from praisonai.cli.output import OutputController, OutputMode
+        
+        controller = OutputController()
+        assert controller.mode == OutputMode.TEXT
+        assert controller.is_json_mode is False
+        assert controller.is_quiet is False
+    
+    def test_json_mode(self):
+        """Test JSON output mode."""
+        from praisonai.cli.output import OutputController, OutputMode
+        
+        controller = OutputController(mode=OutputMode.JSON)
+        assert controller.is_json_mode is True
+    
+    def test_quiet_mode(self):
+        """Test quiet output mode."""
+        from praisonai.cli.output import OutputController, OutputMode
+        
+        controller = OutputController(mode=OutputMode.QUIET)
+        assert controller.is_quiet is True
@@ -0,0 +1,584 @@
+"""
+Unit tests for profiler optimizations module.
+
+Tests Tier 0/1/2 optimization features:
+- Provider caching
+- Lazy imports
+- Client pooling
+- Prewarm manager
+- Lite mode
+- Performance snapshots
+"""
+
+import os
+import tempfile
+import threading
+from pathlib import Path
+from unittest.mock import patch
+
+
+class TestProviderCache:
+    """Tests for ProviderCache (Tier 0)."""
+    
+    def test_singleton_pattern(self):
+        """Test that ProviderCache is a singleton."""
+        from praisonai.cli.features.profiler import ProviderCache
+        
+        cache1 = ProviderCache()
+        cache2 = ProviderCache()
+        assert cache1 is cache2
+    
+    def test_get_set(self):
+        """Test basic get/set operations."""
+        from praisonai.cli.features.profiler import get_provider_cache
+        
+        cache = get_provider_cache()
+        cache.invalidate()  # Clear cache
+        
+        # Miss
+        result = cache.get('test_key')
+        assert result is None
+        
+        # Set
+        cache.set('test_key', {'value': 'test'})
+        
+        # Hit - cache stores as {'value': ..., 'expires': ...}
+        result = cache.get('test_key')
+        assert result is not None
+        assert result['value']['value'] == 'test'
+    
+    def test_invalidate(self):
+        """Test cache invalidation."""
+        from praisonai.cli.features.profiler import get_provider_cache
+        
+        cache = get_provider_cache()
+        cache.set('key1', 'value1')
+        cache.set('key2', 'value2')
+        
+        # Invalidate single key
+        cache.invalidate('key1')
+        assert cache.get('key1') is None
+        assert cache.get('key2') is not None
+        
+        # Invalidate all
+        cache.invalidate()
+        assert cache.get('key2') is None
+    
+    def test_stats(self):
+        """Test cache statistics."""
+        from praisonai.cli.features.profiler import get_provider_cache
+        
+        cache = get_provider_cache()
+        cache.invalidate()
+        
+        # Generate some hits and misses
+        cache.get('miss1')  # miss
+        cache.set('hit1', 'value')
+        cache.get('hit1')  # hit
+        cache.get('miss2')  # miss
+        
+        stats = cache.get_stats()
+        assert 'hits' in stats
+        assert 'misses' in stats
+        assert 'size' in stats
+    
+    def test_thread_safety(self):
+        """Test thread-safe operations."""
+        from praisonai.cli.features.profiler import get_provider_cache
+        
+        cache = get_provider_cache()
+        cache.invalidate()
+        
+        errors = []
+        
+        def worker(thread_id):
+            try:
+                for i in range(100):
+                    key = f"thread_{thread_id}_key_{i}"
+                    cache.set(key, f"value_{i}")
+                    cache.get(key)
+            except Exception as e:
+                errors.append(e)
+        
+        threads = [threading.Thread(target=worker, args=(i,)) for i in range(5)]
+        for t in threads:
+            t.start()
+        for t in threads:
+            t.join()
+        
+        assert len(errors) == 0
+
+
+class TestResolveModelProvider:
+    """Tests for resolve_model_provider (Tier 0)."""
+    
+    def test_openai_models(self):
+        """Test OpenAI model resolution."""
+        from praisonai.cli.features.profiler import resolve_model_provider
+        
+        assert resolve_model_provider('gpt-4o') == ('openai', 'gpt-4o')
+        assert resolve_model_provider('gpt-4o-mini') == ('openai', 'gpt-4o-mini')
+        assert resolve_model_provider('o1-preview') == ('openai', 'o1-preview')
+    
+    def test_anthropic_models(self):
+        """Test Anthropic model resolution."""
+        from praisonai.cli.features.profiler import resolve_model_provider
+        
+        assert resolve_model_provider('claude-3-opus') == ('anthropic', 'claude-3-opus')
+        assert resolve_model_provider('claude-3-sonnet') == ('anthropic', 'claude-3-sonnet')
+    
+    def test_google_models(self):
+        """Test Google model resolution."""
+        from praisonai.cli.features.profiler import resolve_model_provider
+        
+        assert resolve_model_provider('gemini-pro') == ('google', 'gemini-pro')
+        assert resolve_model_provider('gemini-1.5-flash') == ('google', 'gemini-1.5-flash')
+    
+    def test_explicit_provider(self):
+        """Test explicit provider/model format."""
+        from praisonai.cli.features.profiler import resolve_model_provider
+        
+        assert resolve_model_provider('openai/gpt-4') == ('openai', 'gpt-4')
+        assert resolve_model_provider('anthropic/claude-3') == ('anthropic', 'claude-3')
+    
+    def test_caching(self):
+        """Test that results are cached."""
+        from praisonai.cli.features.profiler import resolve_model_provider
+        
+        # Clear cache
+        resolve_model_provider.cache_clear()
+        
+        # First call
+        result1 = resolve_model_provider('gpt-4o')
+        # Second call (should be cached)
+        result2 = resolve_model_provider('gpt-4o')
+        
+        assert result1 == result2
+        
+        # Check cache info
+        info = resolve_model_provider.cache_info()
+        assert info.hits >= 1
+
+
+class TestLazyImporter:
+    """Tests for LazyImporter (Tier 0)."""
+    
+    def test_import_module(self):
+        """Test lazy module import."""
+        from praisonai.cli.features.profiler import LazyImporter
+        
+        LazyImporter.clear()
+        
+        # Import a standard library module
+        json_module = LazyImporter.get('json')
+        assert json_module is not None
+        assert hasattr(json_module, 'dumps')
+    
+    def test_import_attribute(self):
+        """Test lazy attribute import."""
+        from praisonai.cli.features.profiler import LazyImporter
+        
+        LazyImporter.clear()
+        
+        # Import a specific attribute
+        Path = LazyImporter.get('pathlib', 'Path')
+        assert Path is not None
+        assert callable(Path)
+    
+    def test_caching(self):
+        """Test that imports are cached."""
+        from praisonai.cli.features.profiler import LazyImporter
+        
+        LazyImporter.clear()
+        
+        # First import
+        json1 = LazyImporter.get('json')
+        # Second import (should be cached)
+        json2 = LazyImporter.get('json')
+        
+        assert json1 is json2
+
+
+class TestClientPool:
+    """Tests for ClientPool (Tier 1)."""
+    
+    def test_get_or_create(self):
+        """Test client creation and reuse."""
+        from praisonai.cli.features.profiler import ClientPool
+        
+        pool = ClientPool(max_size=5)
+        
+        # Create client
+        client1 = pool.get_or_create('key1', lambda: {'id': 1})
+        assert client1 == {'id': 1}
+        
+        # Reuse client
+        client2 = pool.get_or_create('key1', lambda: {'id': 2})
+        assert client2 == {'id': 1}  # Same as first
+    
+    def test_eviction(self):
+        """Test LRU eviction when pool is full."""
+        from praisonai.cli.features.profiler import ClientPool
+        
+        pool = ClientPool(max_size=2)
+        
+        pool.get_or_create('key1', lambda: 'client1')
+        pool.get_or_create('key2', lambda: 'client2')
+        
+        assert pool.size() == 2
+        
+        # This should evict key1 (oldest)
+        pool.get_or_create('key3', lambda: 'client3')
+        
+        assert pool.size() == 2
+    
+    def test_remove(self):
+        """Test explicit client removal."""
+        from praisonai.cli.features.profiler import ClientPool
+        
+        pool = ClientPool()
+        pool.get_or_create('key1', lambda: 'client1')
+        
+        assert pool.size() == 1
+        
+        pool.remove('key1')
+        
+        assert pool.size() == 0
+    
+    def test_clear(self):
+        """Test clearing all clients."""
+        from praisonai.cli.features.profiler import ClientPool
+        
+        pool = ClientPool()
+        pool.get_or_create('key1', lambda: 'client1')
+        pool.get_or_create('key2', lambda: 'client2')
+        
+        pool.clear()
+        
+        assert pool.size() == 0
+
+
+class TestPrewarmManager:
+    """Tests for PrewarmManager (Tier 1)."""
+    
+    def test_enable_disable(self):
+        """Test enabling and disabling prewarm."""
+        from praisonai.cli.features.profiler import PrewarmManager
+        
+        PrewarmManager.reset()
+        
+        assert not PrewarmManager.is_enabled()
+        
+        PrewarmManager.enable()
+        assert PrewarmManager.is_enabled()
+        
+        PrewarmManager.disable()
+        assert not PrewarmManager.is_enabled()
+    
+    def test_prewarm_disabled(self):
+        """Test that prewarm does nothing when disabled."""
+        from praisonai.cli.features.profiler import PrewarmManager
+        
+        PrewarmManager.reset()
+        
+        # Should not raise even without API key
+        PrewarmManager.prewarm_provider('openai')
+        
+        assert not PrewarmManager.is_prewarmed('openai')
+    
+    def test_reset(self):
+        """Test reset clears state."""
+        from praisonai.cli.features.profiler import PrewarmManager
+        
+        PrewarmManager.enable()
+        PrewarmManager.reset()
+        
+        assert not PrewarmManager.is_enabled()
+
+
+class TestLiteModeConfig:
+    """Tests for LiteModeConfig (Tier 2)."""
+    
+    def test_default_disabled(self):
+        """Test that lite mode is disabled by default."""
+        from praisonai.cli.features.profiler import LiteModeConfig
+        
+        config = LiteModeConfig()
+        
+        assert not config.enabled
+        assert not config.skip_type_validation
+        assert not config.skip_model_validation
+        assert not config.minimal_imports
+    
+    def test_from_env(self):
+        """Test loading config from environment."""
+        from praisonai.cli.features.profiler import LiteModeConfig
+        
+        with patch.dict(os.environ, {
+            'PRAISONAI_LITE_MODE': '1',
+            'PRAISONAI_SKIP_TYPE_VALIDATION': 'true',
+        }):
+            config = LiteModeConfig.from_env()
+            
+            assert config.enabled
+            assert config.skip_type_validation
+            assert not config.skip_model_validation
+    
+    def test_is_lite_mode(self):
+        """Test is_lite_mode helper."""
+        from praisonai.cli.features.profiler import is_lite_mode
+        
+        # Reset global config
+        import praisonai.cli.features.profiler.optimizations as opt
+        opt._lite_mode_config = None
+        
+        with patch.dict(os.environ, {'PRAISONAI_LITE_MODE': ''}, clear=False):
+            # Force reload
+            opt._lite_mode_config = None
+            assert not is_lite_mode()
+
+
+class TestPerfSnapshot:
+    """Tests for PerfSnapshot (Tier 2)."""
+    
+    def test_to_dict(self):
+        """Test snapshot serialization."""
+        from praisonai.cli.features.profiler import PerfSnapshot
+        
+        snapshot = PerfSnapshot(
+            timestamp='2025-01-01T00:00:00Z',
+            name='test',
+            startup_cold_ms=100.0,
+            startup_warm_ms=80.0,
+            import_time_ms=500.0,
+            query_time_ms=2000.0,
+            first_token_ms=300.0,
+        )
+        
+        data = snapshot.to_dict()
+        
+        assert data['name'] == 'test'
+        assert data['startup_cold_ms'] == 100.0
+        assert data['query_time_ms'] == 2000.0
+    
+    def test_from_dict(self):
+        """Test snapshot deserialization."""
+        from praisonai.cli.features.profiler import PerfSnapshot
+        
+        data = {
+            'timestamp': '2025-01-01T00:00:00Z',
+            'name': 'test',
+            'startup_cold_ms': 100.0,
+            'startup_warm_ms': 80.0,
+            'import_time_ms': 500.0,
+            'query_time_ms': 2000.0,
+            'first_token_ms': 300.0,
+        }
+        
+        snapshot = PerfSnapshot.from_dict(data)
+        
+        assert snapshot.name == 'test'
+        assert snapshot.startup_cold_ms == 100.0
+
+
+class TestPerfComparison:
+    """Tests for PerfComparison (Tier 2)."""
+    
+    def test_diff_calculation(self):
+        """Test difference calculations."""
+        from praisonai.cli.features.profiler import PerfSnapshot, PerfComparison
+        
+        baseline = PerfSnapshot(
+            timestamp='2025-01-01T00:00:00Z',
+            name='baseline',
+            startup_cold_ms=100.0,
+            startup_warm_ms=80.0,
+            import_time_ms=500.0,
+            query_time_ms=2000.0,
+            first_token_ms=300.0,
+        )
+        
+        current = PerfSnapshot(
+            timestamp='2025-01-02T00:00:00Z',
+            name='current',
+            startup_cold_ms=110.0,  # 10% slower
+            startup_warm_ms=85.0,
+            import_time_ms=550.0,  # 10% slower
+            query_time_ms=2200.0,  # 10% slower
+            first_token_ms=330.0,
+        )
+        
+        comparison = PerfComparison(baseline=baseline, current=current)
+        
+        assert comparison.startup_cold_diff_ms == 10.0
+        assert comparison.startup_cold_diff_pct == 10.0
+        assert comparison.import_time_diff_ms == 50.0
+        assert comparison.query_time_diff_ms == 200.0
+    
+    def test_regression_detection(self):
+        """Test regression detection."""
+        from praisonai.cli.features.profiler import PerfSnapshot, PerfComparison
+        
+        baseline = PerfSnapshot(
+            timestamp='2025-01-01T00:00:00Z',
+            name='baseline',
+            startup_cold_ms=100.0,
+            startup_warm_ms=80.0,
+            import_time_ms=500.0,
+            query_time_ms=2000.0,
+            first_token_ms=300.0,
+        )
+        
+        # No regression (within 10%)
+        current_ok = PerfSnapshot(
+            timestamp='2025-01-02T00:00:00Z',
+            name='current',
+            startup_cold_ms=105.0,
+            startup_warm_ms=82.0,
+            import_time_ms=520.0,
+            query_time_ms=2100.0,
+            first_token_ms=310.0,
+        )
+        
+        comparison_ok = PerfComparison(baseline=baseline, current=current_ok)
+        assert not comparison_ok.is_regression()
+        
+        # Regression (>10%)
+        current_bad = PerfSnapshot(
+            timestamp='2025-01-02T00:00:00Z',
+            name='current',
+            startup_cold_ms=150.0,  # 50% slower
+            startup_warm_ms=82.0,
+            import_time_ms=520.0,
+            query_time_ms=2100.0,
+            first_token_ms=310.0,
+        )
+        
+        comparison_bad = PerfComparison(baseline=baseline, current=current_bad)
+        assert comparison_bad.is_regression()
+
+
+class TestPerfSnapshotManager:
+    """Tests for PerfSnapshotManager (Tier 2)."""
+    
+    def test_save_and_load(self):
+        """Test saving and loading snapshots."""
+        from praisonai.cli.features.profiler import PerfSnapshot, PerfSnapshotManager
+        
+        with tempfile.TemporaryDirectory() as tmpdir:
+            manager = PerfSnapshotManager(snapshot_dir=Path(tmpdir))
+            
+            snapshot = PerfSnapshot(
+                timestamp='2025-01-01T00:00:00Z',
+                name='test',
+                startup_cold_ms=100.0,
+                startup_warm_ms=80.0,
+                import_time_ms=500.0,
+                query_time_ms=2000.0,
+                first_token_ms=300.0,
+            )
+            
+            path = manager.save(snapshot)
+            assert path.exists()
+            
+            loaded = manager.load('test')
+            assert loaded is not None
+            assert loaded.name == 'test'
+            assert loaded.startup_cold_ms == 100.0
+    
+    def test_baseline(self):
+        """Test baseline save and load."""
+        from praisonai.cli.features.profiler import PerfSnapshot, PerfSnapshotManager
+        
+        with tempfile.TemporaryDirectory() as tmpdir:
+            manager = PerfSnapshotManager(snapshot_dir=Path(tmpdir))
+            
+            snapshot = PerfSnapshot(
+                timestamp='2025-01-01T00:00:00Z',
+                name='original',
+                startup_cold_ms=100.0,
+                startup_warm_ms=80.0,
+                import_time_ms=500.0,
+                query_time_ms=2000.0,
+                first_token_ms=300.0,
+            )
+            
+            manager.save_baseline(snapshot)
+            
+            baseline = manager.load_baseline()
+            assert baseline is not None
+            assert baseline.name == 'baseline'
+    
+    def test_list_snapshots(self):
+        """Test listing snapshots."""
+        from praisonai.cli.features.profiler import PerfSnapshot, PerfSnapshotManager
+        
+        with tempfile.TemporaryDirectory() as tmpdir:
+            manager = PerfSnapshotManager(snapshot_dir=Path(tmpdir))
+            
+            for name in ['snap1', 'snap2', 'snap3']:
+                snapshot = PerfSnapshot(
+                    timestamp='2025-01-01T00:00:00Z',
+                    name=name,
+                    startup_cold_ms=100.0,
+                    startup_warm_ms=80.0,
+                    import_time_ms=500.0,
+                    query_time_ms=2000.0,
+                    first_token_ms=300.0,
+                )
+                manager.save(snapshot)
+            
+            names = manager.list_snapshots()
+            assert 'snap1' in names
+            assert 'snap2' in names
+            assert 'snap3' in names
+
+
+class TestFormatComparisonReport:
+    """Tests for format_comparison_report."""
+    
+    def test_format_report(self):
+        """Test report formatting."""
+        from praisonai.cli.features.profiler import (
+            PerfSnapshot, PerfComparison, format_comparison_report
+        )
+        
+        baseline = PerfSnapshot(
+            timestamp='2025-01-01T00:00:00Z',
+            name='baseline',
+            startup_cold_ms=100.0,
+            startup_warm_ms=80.0,
+            import_time_ms=500.0,
+            query_time_ms=2000.0,
+            first_token_ms=300.0,
+        )
+        
+        current = PerfSnapshot(
+            timestamp='2025-01-02T00:00:00Z',
+            name='current',
+            startup_cold_ms=105.0,
+            startup_warm_ms=82.0,
+            import_time_ms=520.0,
+            query_time_ms=2100.0,
+            first_token_ms=310.0,
+        )
+        
+        comparison = PerfComparison(baseline=baseline, current=current)
+        report = format_comparison_report(comparison)
+        
+        assert 'Performance Comparison Report' in report
+        assert 'baseline' in report
+        assert 'current' in report
+        assert 'No significant regression' in report
+
+
+class TestModuleImports:
+    """Test that all exports are available."""
+    
+    def test_all_exports_available(self):
+        """Test that all __all__ exports are importable."""
+        from praisonai.cli.features.profiler import __all__
+        import praisonai.cli.features.profiler as profiler
+        
+        for name in __all__:
+            assert hasattr(profiler, name), f"Missing export: {name}"
@@ -0,0 +1,228 @@
+"""
+Unit tests for CLI profiler suite module.
+"""
+
+
+class TestScenarioConfig:
+    """Tests for ScenarioConfig dataclass."""
+    
+    def test_default_config(self):
+        """Test default configuration values."""
+        from praisonai.cli.features.profiler import ScenarioConfig
+        
+        config = ScenarioConfig(name="test", prompt="hi")
+        assert config.name == "test"
+        assert config.prompt == "hi"
+        assert config.model is None
+        assert config.stream is False
+        assert config.iterations == 3
+        assert config.warmup == 1
+        assert config.show_files is True
+        assert config.limit == 20
+    
+    def test_custom_config(self):
+        """Test custom configuration."""
+        from praisonai.cli.features.profiler import ScenarioConfig
+        
+        config = ScenarioConfig(
+            name="custom",
+            prompt="test prompt",
+            model="gpt-4o",
+            stream=True,
+            iterations=5,
+        )
+        assert config.name == "custom"
+        assert config.model == "gpt-4o"
+        assert config.stream is True
+        assert config.iterations == 5
+
+
+class TestScenarioResult:
+    """Tests for ScenarioResult dataclass."""
+    
+    def test_empty_result(self):
+        """Test empty result."""
+        from praisonai.cli.features.profiler import ScenarioConfig, ScenarioResult
+        
+        config = ScenarioConfig(name="test", prompt="hi")
+        result = ScenarioResult(name="test", config=config)
+        
+        assert result.name == "test"
+        assert len(result.results) == 0
+        assert result.import_times == []
+        assert result.total_times == []
+    
+    def test_get_stats_empty(self):
+        """Test stats with empty values."""
+        from praisonai.cli.features.profiler import ScenarioConfig, ScenarioResult
+        
+        config = ScenarioConfig(name="test", prompt="hi")
+        result = ScenarioResult(name="test", config=config)
+        
+        stats = result.get_stats([])
+        assert stats["mean"] == 0
+        assert stats["min"] == 0
+        assert stats["max"] == 0
+    
+    def test_get_stats_single_value(self):
+        """Test stats with single value."""
+        from praisonai.cli.features.profiler import ScenarioConfig, ScenarioResult
+        
+        config = ScenarioConfig(name="test", prompt="hi")
+        result = ScenarioResult(name="test", config=config)
+        
+        stats = result.get_stats([100.0])
+        assert stats["mean"] == 100.0
+        assert stats["min"] == 100.0
+        assert stats["max"] == 100.0
+        assert stats["stdev"] == 0
+    
+    def test_get_stats_multiple_values(self):
+        """Test stats with multiple values."""
+        from praisonai.cli.features.profiler import ScenarioConfig, ScenarioResult
+        
+        config = ScenarioConfig(name="test", prompt="hi")
+        result = ScenarioResult(name="test", config=config)
+        
+        stats = result.get_stats([100.0, 200.0, 300.0])
+        assert stats["mean"] == 200.0
+        assert stats["min"] == 100.0
+        assert stats["max"] == 300.0
+        assert stats["median"] == 200.0
+    
+    def test_to_dict(self):
+        """Test conversion to dictionary."""
+        from praisonai.cli.features.profiler import ScenarioConfig, ScenarioResult
+        
+        config = ScenarioConfig(name="test", prompt="hi", iterations=2)
+        result = ScenarioResult(name="test", config=config)
+        
+        data = result.to_dict()
+        assert data["name"] == "test"
+        assert "config" in data
+        assert data["config"]["prompt"] == "hi"
+        assert "import_time_stats" in data
+        assert "total_time_stats" in data
+
+
+class TestSuiteResult:
+    """Tests for SuiteResult dataclass."""
+    
+    def test_default_result(self):
+        """Test default result."""
+        from praisonai.cli.features.profiler import SuiteResult
+        
+        result = SuiteResult()
+        assert len(result.scenarios) == 0
+        assert result.startup_cold_ms == 0.0
+        assert result.startup_warm_ms == 0.0
+        assert len(result.import_analysis) == 0
+        assert result.timestamp.endswith("Z")  # ISO format with Z suffix
+    
+    def test_to_dict(self):
+        """Test conversion to dictionary."""
+        from praisonai.cli.features.profiler import SuiteResult
+        
+        result = SuiteResult(
+            startup_cold_ms=100.0,
+            startup_warm_ms=50.0,
+        )
+        
+        data = result.to_dict()
+        assert "timestamp" in data
+        assert "metadata" in data
+        assert data["startup"]["cold_ms"] == 100.0
+        assert data["startup"]["warm_ms"] == 50.0
+
+
+class TestProfileSuiteRunner:
+    """Tests for ProfileSuiteRunner class."""
+    
+    def test_init_default(self):
+        """Test initialization with defaults."""
+        from praisonai.cli.features.profiler import ProfileSuiteRunner
+        
+        runner = ProfileSuiteRunner()
+        assert len(runner.scenarios) == 4  # DEFAULT_SCENARIOS
+        assert runner.verbose is False
+    
+    def test_init_custom_scenarios(self):
+        """Test initialization with custom scenarios."""
+        from praisonai.cli.features.profiler import ProfileSuiteRunner, ScenarioConfig
+        
+        scenarios = [
+            ScenarioConfig(name="custom1", prompt="test1"),
+            ScenarioConfig(name="custom2", prompt="test2"),
+        ]
+        
+        runner = ProfileSuiteRunner(scenarios=scenarios)
+        assert len(runner.scenarios) == 2
+        assert runner.scenarios[0].name == "custom1"
+    
+    def test_collect_metadata(self):
+        """Test metadata collection."""
+        from praisonai.cli.features.profiler import ProfileSuiteRunner
+        
+        runner = ProfileSuiteRunner()
+        metadata = runner._collect_metadata()
+        
+        assert "python_version" in metadata
+        assert "platform" in metadata
+        assert "praisonai_version" in metadata
+        assert "timestamp" in metadata
+    
+    def test_measure_startup(self):
+        """Test startup measurement."""
+        from praisonai.cli.features.profiler import ProfileSuiteRunner
+        
+        runner = ProfileSuiteRunner()
+        cold, warm = runner._measure_startup()
+        
+        # Both should be positive numbers
+        assert cold > 0
+        assert warm > 0
+    
+    def test_analyze_imports(self):
+        """Test import analysis."""
+        from praisonai.cli.features.profiler import ProfileSuiteRunner
+        
+        runner = ProfileSuiteRunner()
+        imports = runner._analyze_imports()
+        
+        # Should return a list (may be empty if subprocess fails)
+        assert isinstance(imports, list)
+        
+        # If we got results, check structure
+        if imports:
+            assert "module" in imports[0]
+            assert "cumulative_ms" in imports[0]
+
+
+class TestProfilerModuleImports:
+    """Tests for profiler module imports."""
+    
+    def test_all_exports_available(self):
+        """Test that all exports are available."""
+        from praisonai.cli.features.profiler import (
+            ProfilerConfig,
+            ProfilerResult,
+            QueryProfiler,
+            run_profiled_query,
+            format_profile_report,
+            ScenarioConfig,
+            ScenarioResult,
+            SuiteResult,
+            ProfileSuiteRunner,
+            run_profile_suite,
+        )
+        
+        assert ProfilerConfig is not None
+        assert ProfilerResult is not None
+        assert QueryProfiler is not None
+        assert run_profiled_query is not None
+        assert format_profile_report is not None
+        assert ScenarioConfig is not None
+        assert ScenarioResult is not None
+        assert SuiteResult is not None
+        assert ProfileSuiteRunner is not None
+        assert run_profile_suite is not None
@@ -0,0 +1,630 @@
+"""
+Unit tests for the Unified Profiling Architecture.
+
+Tests the core invariants:
+1. Profiled execution produces identical output to non-profiled execution
+2. Schema equivalence across invocation methods
+3. Overhead bounds
+4. Profile data bounds
+"""
+
+import pytest
+from unittest.mock import patch
+
+
+class TestExecutionRequest:
+    """Tests for ExecutionRequest dataclass."""
+    
+    def test_create_basic_request(self):
+        """Test creating a basic execution request."""
+        from praisonai.cli.execution import ExecutionRequest
+        
+        req = ExecutionRequest(prompt="What is 2+2?")
+        
+        assert req.prompt == "What is 2+2?"
+        assert req.agent_name == "Assistant"
+        assert req.model is None
+        assert req.stream is False
+        assert req.tools == ()
+    
+    def test_request_immutability(self):
+        """Test that ExecutionRequest is immutable (frozen)."""
+        from praisonai.cli.execution import ExecutionRequest
+        
+        req = ExecutionRequest(prompt="test")
+        
+        with pytest.raises(AttributeError):
+            req.prompt = "modified"
+    
+    def test_request_with_model(self):
+        """Test creating request with model."""
+        from praisonai.cli.execution import ExecutionRequest
+        
+        req = ExecutionRequest(prompt="test", model="gpt-4")
+        new_req = req.with_model("gpt-3.5-turbo")
+        
+        assert req.model == "gpt-4"
+        assert new_req.model == "gpt-3.5-turbo"
+        assert req.prompt == new_req.prompt
+    
+    def test_request_validation(self):
+        """Test that empty prompt raises error."""
+        from praisonai.cli.execution import ExecutionRequest
+        
+        with pytest.raises(ValueError):
+            ExecutionRequest(prompt="")
+    
+    def test_request_to_dict(self):
+        """Test serialization to dict."""
+        from praisonai.cli.execution import ExecutionRequest
+        
+        req = ExecutionRequest(
+            prompt="test",
+            agent_name="TestAgent",
+            model="gpt-4",
+        )
+        
+        d = req.to_dict()
+        
+        assert d["prompt"] == "test"
+        assert d["agent_name"] == "TestAgent"
+        assert d["model"] == "gpt-4"
+        assert d["tools_count"] == 0
+
+
+class TestExecutionResult:
+    """Tests for ExecutionResult dataclass."""
+    
+    def test_create_basic_result(self):
+        """Test creating a basic execution result."""
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        result = ExecutionResult(output="four")
+        
+        assert result.output == "four"
+        assert result.success is True
+        assert result.error is None
+    
+    def test_result_from_error(self):
+        """Test creating result from error."""
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        result = ExecutionResult.from_error("Something went wrong")
+        
+        assert result.output == ""
+        assert result.success is False
+        assert result.error == "Something went wrong"
+    
+    def test_result_duration(self):
+        """Test duration calculation."""
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        result = ExecutionResult(
+            output="test",
+            start_time=1000.0,
+            end_time=1002.5,
+        )
+        
+        assert result.duration_ms == 2500.0
+
+
+class TestProfilerConfig:
+    """Tests for ProfilerConfig."""
+    
+    def test_default_config(self):
+        """Test default profiler configuration."""
+        from praisonai.cli.execution.profiler import ProfilerConfig
+        
+        config = ProfilerConfig()
+        
+        assert config.layer == 1
+        assert config.limit == 30
+        assert config.track_network is False
+    
+    def test_from_flags_basic(self):
+        """Test creating config from basic --profile flag."""
+        from praisonai.cli.execution.profiler import ProfilerConfig
+        
+        config = ProfilerConfig.from_flags(profile=True)
+        
+        assert config.layer == 1
+    
+    def test_from_flags_deep(self):
+        """Test creating config from --profile --deep flags."""
+        from praisonai.cli.execution.profiler import ProfilerConfig
+        
+        config = ProfilerConfig.from_flags(profile=True, deep=True)
+        
+        assert config.layer == 2
+        assert config.show_callers is True
+        assert config.show_callees is True
+    
+    def test_from_flags_network(self):
+        """Test creating config with network tracking."""
+        from praisonai.cli.execution.profiler import ProfilerConfig
+        
+        config = ProfilerConfig.from_flags(profile=True, network=True)
+        
+        assert config.track_network is True
+
+
+class TestTimingBreakdown:
+    """Tests for TimingBreakdown."""
+    
+    def test_timing_to_dict(self):
+        """Test timing serialization."""
+        from praisonai.cli.execution.profiler import TimingBreakdown
+        
+        timing = TimingBreakdown(
+            total_ms=1000.0,
+            imports_ms=500.0,
+            agent_init_ms=10.0,
+            execution_ms=490.0,
+        )
+        
+        d = timing.to_dict()
+        
+        assert d["total_ms"] == 1000.0
+        assert d["imports_ms"] == 500.0
+        assert d["agent_init_ms"] == 10.0
+        assert d["execution_ms"] == 490.0
+
+
+class TestProfileReport:
+    """Tests for ProfileReport."""
+    
+    def test_report_schema_version(self):
+        """Test that schema version is set."""
+        from praisonai.cli.execution.profiler import ProfileReport, SCHEMA_VERSION
+        
+        report = ProfileReport(run_id="test123")
+        
+        assert report.schema_version == SCHEMA_VERSION
+        assert report.schema_version == "1.0"
+    
+    def test_report_timestamp(self):
+        """Test that timestamp is auto-generated."""
+        from praisonai.cli.execution.profiler import ProfileReport
+        
+        report = ProfileReport(run_id="test123")
+        
+        assert report.timestamp is not None
+        assert "T" in report.timestamp  # ISO 8601 format
+    
+    def test_report_to_dict(self):
+        """Test report serialization."""
+        from praisonai.cli.execution.profiler import (
+            ProfileReport, TimingBreakdown, InvocationInfo
+        )
+        
+        report = ProfileReport(
+            run_id="test123",
+            timing=TimingBreakdown(total_ms=100),
+            invocation=InvocationInfo(method="test"),
+        )
+        
+        d = report.to_dict()
+        
+        assert d["schema_version"] == "1.0"
+        assert d["run_id"] == "test123"
+        assert "timing" in d
+        assert "invocation" in d
+    
+    def test_report_to_json(self):
+        """Test JSON serialization."""
+        from praisonai.cli.execution.profiler import ProfileReport
+        import json
+        
+        report = ProfileReport(run_id="test123")
+        
+        json_str = report.to_json()
+        parsed = json.loads(json_str)
+        
+        assert parsed["run_id"] == "test123"
+    
+    def test_report_to_text(self):
+        """Test text formatting."""
+        from praisonai.cli.execution.profiler import ProfileReport
+        
+        report = ProfileReport(run_id="test123")
+        
+        text = report.to_text()
+        
+        assert "test123" in text
+        assert "Timing Breakdown" in text
+
+
+class TestProfilerBounds:
+    """Tests for profile data bounds."""
+    
+    def test_max_function_stats(self):
+        """Test that function stats are bounded."""
+        from praisonai.cli.execution.profiler import MAX_FUNCTION_STATS
+        
+        assert MAX_FUNCTION_STATS == 1000
+    
+    def test_max_call_graph_edges(self):
+        """Test that call graph edges are bounded."""
+        from praisonai.cli.execution.profiler import MAX_CALL_GRAPH_EDGES
+        
+        assert MAX_CALL_GRAPH_EDGES == 5000
+    
+    def test_max_network_requests(self):
+        """Test that network requests are bounded."""
+        from praisonai.cli.execution.profiler import MAX_NETWORK_REQUESTS
+        
+        assert MAX_NETWORK_REQUESTS == 100
+
+
+class TestSchemaStability:
+    """Tests for schema stability guarantees."""
+    
+    def test_mandatory_fields_present(self):
+        """Test that all mandatory fields are present in report."""
+        from praisonai.cli.execution.profiler import ProfileReport
+        
+        report = ProfileReport(run_id="test")
+        d = report.to_dict()
+        
+        # Mandatory fields
+        assert "schema_version" in d
+        assert "run_id" in d
+        assert "timestamp" in d
+        assert "invocation" in d
+        assert "timing" in d
+        assert "response_preview" in d
+    
+    def test_optional_fields_nullable(self):
+        """Test that optional fields can be None."""
+        from praisonai.cli.execution.profiler import ProfileReport
+        
+        report = ProfileReport(run_id="test")
+        d = report.to_dict()
+        
+        # Optional fields should not be present if None
+        assert "functions" not in d or d["functions"] is None
+        assert "call_graph" not in d or d["call_graph"] is None
+        assert "network" not in d or d["network"] is None
+
+
+class TestInvocationInfo:
+    """Tests for InvocationInfo."""
+    
+    def test_version_auto_populated(self):
+        """Test that versions are auto-populated."""
+        from praisonai.cli.execution.profiler import InvocationInfo
+        
+        info = InvocationInfo(method="test")
+        
+        assert info.python_version is not None
+        assert len(info.python_version) > 0
+
+
+class TestCallGraph:
+    """Tests for CallGraph."""
+    
+    def test_edges_property(self):
+        """Test edges property."""
+        from praisonai.cli.execution.profiler import CallGraph
+        
+        graph = CallGraph(
+            callers={"func_b": ["func_a"]},
+            callees={"func_a": ["func_b"]},
+        )
+        
+        edges = graph.edges
+        
+        assert len(edges) == 1
+        assert ("func_a", "func_b") in edges
+
+
+class TestRequestTiming:
+    """Tests for RequestTiming."""
+    
+    def test_ttfb_calculation(self):
+        """Test time to first byte calculation."""
+        from praisonai.cli.execution.profiler import RequestTiming
+        
+        timing = RequestTiming(
+            url="https://api.openai.com",
+            method="POST",
+            start_time=1000.0,
+            first_byte_time=1001.5,
+            end_time=1002.0,
+            status_code=200,
+        )
+        
+        assert timing.ttfb_ms == 1500.0
+        assert timing.total_ms == 2000.0
+
+
+# Integration tests (require mocking)
+class TestProfilerIntegration:
+    """Integration tests for Profiler class."""
+    
+    def test_profiler_creates_report(self):
+        """Test that profiler creates a valid report."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        # Mock at the profiler level by patching the imported _execute_core
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="four",
+                run_id="test123",
+            )
+            
+            config = ProfilerConfig(layer=0)  # Minimal profiling
+            profiler = Profiler(config)
+            
+            req = ExecutionRequest(prompt="What is 2+2?")
+            result, report = profiler.profile_sync(req)
+            
+            assert result.output == "four"
+            assert report.schema_version == "1.0"
+            assert report.timing.total_ms > 0
+    
+    def test_output_equivalence_profiled_vs_non_profiled(self):
+        """INVARIANT 1: Profiled execution produces identical output."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        expected_output = "test_output_value"
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output=expected_output,
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            
+            # Layer 0 (minimal)
+            config0 = ProfilerConfig(layer=0)
+            result0, _ = Profiler(config0).profile_sync(req)
+            
+            # Layer 1 (basic)
+            config1 = ProfilerConfig(layer=1)
+            result1, _ = Profiler(config1).profile_sync(req)
+            
+            # Layer 2 (deep)
+            config2 = ProfilerConfig(layer=2)
+            result2, _ = Profiler(config2).profile_sync(req)
+            
+            # All outputs must be identical
+            assert result0.output == expected_output
+            assert result1.output == expected_output
+            assert result2.output == expected_output
+    
+    def test_schema_equivalence_across_invocation_methods(self):
+        """INVARIANT 2: Schema equivalence across invocation methods."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            config = ProfilerConfig(layer=1)
+            
+            # CLI direct
+            _, report1 = Profiler(config).profile_sync(req, invocation_method="cli_direct")
+            
+            # Profile command
+            _, report2 = Profiler(config).profile_sync(req, invocation_method="profile_command")
+            
+            # Schema must be identical
+            dict1 = report1.to_dict()
+            dict2 = report2.to_dict()
+            
+            assert dict1["schema_version"] == dict2["schema_version"]
+            assert set(dict1["timing"].keys()) == set(dict2["timing"].keys())
+            assert set(dict1["invocation"].keys()) == set(dict2["invocation"].keys())
+    
+    def test_data_bounds_function_stats(self):
+        """INVARIANT 4: Function stats are bounded."""
+        from praisonai.cli.execution.profiler import (
+            Profiler, ProfilerConfig, MAX_FUNCTION_STATS
+        )
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            config = ProfilerConfig(layer=1, limit=50)
+            
+            _, report = Profiler(config).profile_sync(req)
+            
+            # Functions should be bounded by limit
+            if report.functions:
+                assert len(report.functions) <= config.limit
+                assert len(report.functions) <= MAX_FUNCTION_STATS
+    
+    def test_layer_0_minimal_overhead(self):
+        """INVARIANT 3: Layer 0 has minimal overhead (<1ms)."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        import time
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            # Fast mock execution
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            config = ProfilerConfig(layer=0)
+            
+            start = time.perf_counter()
+            Profiler(config).profile_sync(req)
+            elapsed = (time.perf_counter() - start) * 1000
+            
+            # Layer 0 overhead should be minimal (allow some margin for test environment)
+            # The actual profiling overhead (excluding execution) should be < 1ms
+            # We can't measure this precisely with mocks, but we verify it runs fast
+            assert elapsed < 100  # Very generous bound for test environment
+    
+    def test_ttfr_is_populated(self):
+        """Test that Time to First Response (TTFR) is populated and > 0."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            config = ProfilerConfig(layer=1)
+            
+            _, report = Profiler(config).profile_sync(req)
+            
+            # TTFR must be populated
+            assert report.timing.time_to_first_response_ms > 0
+            assert report.timing.first_output_ms >= 0
+    
+    def test_first_output_less_than_total(self):
+        """Test that first_output_ms <= total_ms."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            config = ProfilerConfig(layer=1)
+            
+            _, report = Profiler(config).profile_sync(req)
+            
+            # first_output must be <= total
+            assert report.timing.first_output_ms <= report.timing.total_ms
+    
+    def test_timeline_phases_ordered(self):
+        """Test that timeline phases are correctly ordered."""
+        from praisonai.cli.execution.profiler import TimingBreakdown
+        
+        timing = TimingBreakdown(
+            total_ms=1000.0,
+            imports_ms=200.0,
+            agent_init_ms=50.0,
+            execution_ms=700.0,
+        )
+        
+        timeline = timing.to_timeline()
+        
+        # Timeline should have entries
+        assert len(timeline) > 0
+        
+        # Each entry should have (name, start, duration)
+        for entry in timeline:
+            assert len(entry) == 3
+            name, start, duration = entry
+            assert isinstance(name, str)
+            assert isinstance(start, float)
+            assert isinstance(duration, float)
+    
+    def test_decision_trace_in_deep_profile(self):
+        """Test that decision trace is present in deep profile (layer 2)."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test", model="gpt-4")
+            config = ProfilerConfig(layer=2)  # Deep profile
+            
+            _, report = Profiler(config).profile_sync(req)
+            
+            # Decision trace must be present in deep profile
+            assert report.decision_trace is not None
+            assert report.decision_trace.model_selected == "gpt-4"
+            assert report.decision_trace.profile_layer == 2
+    
+    def test_module_breakdown_in_deep_profile(self):
+        """Test that module breakdown is present in deep profile (layer 2)."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            config = ProfilerConfig(layer=2)  # Deep profile
+            
+            _, report = Profiler(config).profile_sync(req)
+            
+            # Module breakdown must be present in deep profile
+            assert report.module_breakdown is not None
+    
+    def test_text_output_includes_timeline(self):
+        """Test that text output includes timeline section."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            config = ProfilerConfig(layer=1)
+            
+            _, report = Profiler(config).profile_sync(req)
+            text = report.to_text()
+            
+            # Text output must include timeline and TTFR
+            assert "Execution Timeline" in text
+            assert "Time to First Response" in text
+    
+    def test_json_output_includes_ttfr(self):
+        """Test that JSON output includes time_to_first_response_ms."""
+        from praisonai.cli.execution.profiler import Profiler, ProfilerConfig
+        from praisonai.cli.execution.request import ExecutionRequest
+        from praisonai.cli.execution.result import ExecutionResult
+        import json
+        
+        with patch('praisonai.cli.execution.profiler._execute_core') as mock_exec:
+            mock_exec.return_value = ExecutionResult(
+                output="test",
+                run_id="test123",
+            )
+            
+            req = ExecutionRequest(prompt="test")
+            config = ProfilerConfig(layer=1)
+            
+            _, report = Profiler(config).profile_sync(req)
+            data = json.loads(report.to_json())
+            
+            # JSON must include TTFR
+            assert "time_to_first_response_ms" in data["timing"]
+            assert data["timing"]["time_to_first_response_ms"] > 0
@@ -0,0 +1,330 @@
+"""
+TDD Test Suite for PraisonAI TUI Widgets.
+
+Tests for:
+- ComposerWidget (Enter to send, Shift+Enter for newline)
+- HelpScreen (F1 key)
+- Keyboard shortcuts (Ctrl+Q, Ctrl+C, Ctrl+L)
+- Queue panel (edit/delete)
+- Tool panel (display tools)
+"""
+
+import pytest
+from unittest.mock import MagicMock, patch, AsyncMock
+
+# Check if textual is available
+try:
+    from textual.pilot import Pilot
+    from textual.app import App
+    TEXTUAL_AVAILABLE = True
+except ImportError:
+    TEXTUAL_AVAILABLE = False
+    pytest.skip("Textual not available", allow_module_level=True)
+
+
+class TestComposerWidget:
+    """Tests for ComposerWidget."""
+    
+    @pytest.mark.asyncio
+    async def test_submit_text_area_enter_triggers_submit(self):
+        """Test that Enter key triggers submit."""
+        from praisonai.cli.features.tui.widgets.composer import SubmitTextArea
+        
+        # Create a mock app context
+        text_area = SubmitTextArea("")
+        
+        # Verify the class exists and has SubmitRequested message
+        assert hasattr(SubmitTextArea, 'SubmitRequested')
+    
+    @pytest.mark.asyncio
+    async def test_composer_widget_has_correct_bindings(self):
+        """Test ComposerWidget has correct key bindings."""
+        from praisonai.cli.features.tui.widgets.composer import ComposerWidget
+        
+        # Check bindings include enter for submit
+        binding_keys = [b.key for b in ComposerWidget.BINDINGS]
+        assert "enter" in binding_keys
+        assert "escape" in binding_keys
+
+
+class TestHelpScreen:
+    """Tests for HelpScreen."""
+    
+    def test_help_screen_exists(self):
+        """Test HelpScreen can be imported."""
+        from praisonai.cli.features.tui.screens.help import HelpScreen
+        assert HelpScreen is not None
+    
+    def test_help_screen_has_dismiss_binding(self):
+        """Test HelpScreen has escape binding to dismiss."""
+        from praisonai.cli.features.tui.screens.help import HelpScreen
+        
+        binding_keys = [b.key for b in HelpScreen.BINDINGS]
+        assert "escape" in binding_keys
+        assert "q" in binding_keys
+
+
+class TestTUIApp:
+    """Tests for TUIApp."""
+    
+    def test_tui_app_has_help_screen_registered(self):
+        """Test TUIApp has help screen in SCREENS."""
+        from praisonai.cli.features.tui.app import TUIApp
+        
+        assert "help" in TUIApp.SCREENS
+        assert "main" in TUIApp.SCREENS
+        assert "settings" in TUIApp.SCREENS
+        assert "queue" in TUIApp.SCREENS
+    
+    def test_tui_app_has_ctrl_bindings(self):
+        """Test TUIApp has Ctrl key bindings."""
+        from praisonai.cli.features.tui.app import TUIApp
+        
+        binding_keys = [b.key for b in TUIApp.BINDINGS]
+        assert "ctrl+q" in binding_keys
+        assert "ctrl+c" in binding_keys
+
+
+class TestMainScreen:
+    """Tests for MainScreen."""
+    
+    def test_main_screen_has_safe_default_bindings(self):
+        """Test MainScreen has safe default bindings (no Ctrl/Fn by default)."""
+        from praisonai.cli.features.tui.screens.main import MainScreen
+        
+        binding_keys = [b.key for b in MainScreen.BINDINGS]
+        # Safe defaults - single char shortcuts
+        assert "q" in binding_keys  # quit
+        assert "colon" in binding_keys  # command mode
+        assert "escape" in binding_keys  # cancel
+        # Ctrl keys should NOT be in default bindings
+        assert "ctrl+q" not in binding_keys
+        assert "ctrl+c" not in binding_keys
+        # Function keys should NOT be in default bindings
+        assert "f1" not in binding_keys
+
+
+class TestTUIConfig:
+    """Tests for TUIConfig."""
+    
+    def test_default_config_disables_ctrl_keys(self):
+        """Test default config has Ctrl keys disabled."""
+        from praisonai.cli.features.tui.config import TUIConfig
+        
+        config = TUIConfig()
+        assert config.enable_ctrl_keys is False
+        assert config.enable_fn_keys is False
+        assert config.enable_single_char is True
+        assert config.leader_key == ":"
+    
+    def test_command_parsing(self):
+        """Test command mode parsing."""
+        from praisonai.cli.features.tui.config import get_command
+        
+        assert get_command("quit") == "quit"
+        assert get_command("q") == "quit"
+        assert get_command("exit") == "quit"
+        assert get_command("clear") == "clear"
+        assert get_command("cl") == "clear"
+        assert get_command("invalid") is None
+
+
+class TestQueuePanel:
+    """Tests for QueuePanelWidget."""
+    
+    def test_queue_panel_has_delete_message(self):
+        """Test QueuePanelWidget has DeleteRequested message."""
+        from praisonai.cli.features.tui.widgets.queue_panel import QueuePanelWidget
+        
+        assert hasattr(QueuePanelWidget, 'DeleteRequested')
+        assert hasattr(QueuePanelWidget, 'EditRequested')
+    
+    def test_queue_panel_has_delete_method(self):
+        """Test QueuePanelWidget has delete_selected method."""
+        from praisonai.cli.features.tui.widgets.queue_panel import QueuePanelWidget
+        
+        assert hasattr(QueuePanelWidget, 'delete_selected')
+        assert hasattr(QueuePanelWidget, 'edit_selected')
+
+
+class TestToolPanel:
+    """Tests for ToolPanelWidget."""
+    
+    def test_tool_panel_has_set_available_tools(self):
+        """Test ToolPanelWidget has set_available_tools method."""
+        from praisonai.cli.features.tui.widgets.tool_panel import ToolPanelWidget
+        
+        assert hasattr(ToolPanelWidget, 'set_available_tools')
+
+
+class TestInteractiveTools:
+    """Tests for interactive tools loading."""
+    
+    def test_get_interactive_tools_exists(self):
+        """Test get_interactive_tools function exists."""
+        from praisonai.cli.features.interactive_tools import get_interactive_tools
+        
+        assert callable(get_interactive_tools)
+    
+    def test_tool_config_exists(self):
+        """Test ToolConfig class exists."""
+        from praisonai.cli.features.interactive_tools import ToolConfig
+        
+        assert ToolConfig is not None
+
+
+class TestCommandPopup:
+    """Tests for CommandPopupWidget."""
+    
+    def test_command_popup_exists(self):
+        """Test CommandPopupWidget can be imported."""
+        from praisonai.cli.features.tui.widgets.command_popup import CommandPopupWidget
+        assert CommandPopupWidget is not None
+    
+    def test_command_info_dataclass(self):
+        """Test CommandInfo dataclass works correctly."""
+        from praisonai.cli.features.tui.widgets.command_popup import CommandInfo
+        
+        cmd = CommandInfo(
+            name="test",
+            description="Test command",
+            aliases=["t", "tst"]
+        )
+        assert cmd.name == "test"
+        assert cmd.description == "Test command"
+        assert cmd.aliases == ["t", "tst"]
+        assert cmd.display_name == "test (t, tst)"
+    
+    def test_get_default_commands(self):
+        """Test get_default_commands returns command list."""
+        from praisonai.cli.features.tui.widgets.command_popup import get_default_commands
+        
+        commands = get_default_commands()
+        assert len(commands) > 0
+        
+        # Check for essential commands
+        command_names = [c.name for c in commands]
+        assert "help" in command_names
+        assert "exit" in command_names
+        assert "clear" in command_names
+        assert "model" in command_names
+    
+    def test_command_popup_has_messages(self):
+        """Test CommandPopupWidget has required message classes."""
+        from praisonai.cli.features.tui.widgets.command_popup import CommandPopupWidget
+        
+        assert hasattr(CommandPopupWidget, 'CommandSelected')
+        assert hasattr(CommandPopupWidget, 'Dismissed')
+
+
+class TestMainScreenBackslash:
+    """Tests for MainScreen backslash command support."""
+    
+    def test_main_screen_has_backslash_binding(self):
+        """Test MainScreen has backslash binding for commands."""
+        from praisonai.cli.features.tui.screens.main import MainScreen
+        
+        binding_keys = [b.key for b in MainScreen.BINDINGS]
+        assert "backslash" in binding_keys
+    
+    def test_main_screen_has_show_commands_action(self):
+        """Test MainScreen has action_show_commands method."""
+        from praisonai.cli.features.tui.screens.main import MainScreen
+        
+        assert hasattr(MainScreen, 'action_show_commands')
+    
+    def test_main_screen_has_command_popup_handlers(self):
+        """Test MainScreen has command popup event handlers."""
+        from praisonai.cli.features.tui.screens.main import MainScreen
+        
+        assert hasattr(MainScreen, 'on_command_popup_widget_command_selected')
+        assert hasattr(MainScreen, 'on_command_popup_widget_dismissed')
+
+
+class TestAutoTriggerEvents:
+    """Tests for auto-trigger events (backslash, @, /)."""
+    
+    def test_composer_has_backslash_typed_message(self):
+        """Test ComposerWidget has BackslashTyped message."""
+        from praisonai.cli.features.tui.widgets.composer import ComposerWidget
+        
+        assert hasattr(ComposerWidget, 'BackslashTyped')
+    
+    def test_composer_has_at_typed_message(self):
+        """Test ComposerWidget has AtTyped message for @ file autocomplete."""
+        from praisonai.cli.features.tui.widgets.composer import ComposerWidget
+        
+        assert hasattr(ComposerWidget, 'AtTyped')
+    
+    def test_composer_has_slash_typed_message(self):
+        """Test ComposerWidget has SlashTyped message for / commands."""
+        from praisonai.cli.features.tui.widgets.composer import ComposerWidget
+        
+        assert hasattr(ComposerWidget, 'SlashTyped')
+    
+    def test_main_screen_has_auto_trigger_handlers(self):
+        """Test MainScreen has auto-trigger event handlers."""
+        from praisonai.cli.features.tui.screens.main import MainScreen
+        
+        assert hasattr(MainScreen, 'on_composer_widget_backslash_typed')
+        assert hasattr(MainScreen, 'on_composer_widget_at_typed')
+        assert hasattr(MainScreen, 'on_composer_widget_slash_typed')
+
+
+class TestFilePopup:
+    """Tests for FilePopupWidget."""
+    
+    def test_file_popup_exists(self):
+        """Test FilePopupWidget can be imported."""
+        from praisonai.cli.features.tui.widgets.file_popup import FilePopupWidget
+        assert FilePopupWidget is not None
+    
+    def test_file_popup_has_messages(self):
+        """Test FilePopupWidget has required message classes."""
+        from praisonai.cli.features.tui.widgets.file_popup import FilePopupWidget
+        
+        assert hasattr(FilePopupWidget, 'FileSelected')
+        assert hasattr(FilePopupWidget, 'Dismissed')
+    
+    def test_file_info_dataclass(self):
+        """Test FileInfo dataclass works correctly."""
+        from praisonai.cli.features.tui.widgets.file_popup import FileInfo
+        
+        file_info = FileInfo(path="test.py", file_type="file")
+        assert file_info.path == "test.py"
+        assert file_info.file_type == "file"
+        assert file_info.icon == "📄 "
+        
+        dir_info = FileInfo(path="src/", file_type="directory")
+        assert dir_info.icon == "📁 "
+    
+    def test_main_screen_has_file_popup_handlers(self):
+        """Test MainScreen has file popup event handlers."""
+        from praisonai.cli.features.tui.screens.main import MainScreen
+        
+        assert hasattr(MainScreen, 'on_file_popup_widget_file_selected')
+        assert hasattr(MainScreen, 'on_file_popup_widget_dismissed')
+        assert hasattr(MainScreen, '_show_file_popup')
+        assert hasattr(MainScreen, '_dismiss_file_popup')
+
+
+class TestQueueLiveUpdates:
+    """Tests for queue panel live updates."""
+    
+    def test_app_has_update_queue_panel_live(self):
+        """Test TUIApp has _update_queue_panel_live method."""
+        from praisonai.cli.features.tui.app import TUIApp
+        
+        assert hasattr(TUIApp, '_update_queue_panel_live')
+    
+    def test_queue_panel_has_reactive_counts(self):
+        """Test QueuePanelWidget has reactive count properties."""
+        from praisonai.cli.features.tui.widgets.queue_panel import QueuePanelWidget
+        
+        # Check reactive properties exist
+        assert hasattr(QueuePanelWidget, 'queued_count')
+        assert hasattr(QueuePanelWidget, 'running_count')
+
+
+if __name__ == "__main__":
+    pytest.main([__file__, "-v"])
@@ -4276,7 +4276,7 @@ requires-dist = [
    { name = "praisonai-tools", marker = "extra == 'autogen'", specifier = ">=0.1.0" },
    { name = "praisonai-tools", marker = "extra == 'autogen-v4'", specifier = ">=0.1.0" },
    { name = "praisonai-tools", marker = "extra == 'crewai'", specifier = ">=0.1.0" },
-    { name = "praisonaiagents", specifier = ">=0.5.3" },
+    { name = "praisonaiagents", specifier = ">=0.6.0" },
    { name = "pyautogen", marker = "extra == 'autogen'", specifier = "==0.2.29" },
    { name = "pydantic", marker = "extra == 'chat'", specifier = "<=2.10.1" },
    { name = "pydantic", marker = "extra == 'code'", specifier = "<=2.10.1" },
@@ -4324,7 +4324,7 @@ wheels = [

 [[package]]
 name = "praisonaiagents"
-version = "0.5.3"
+version = "0.6.0"
 source = { registry = "https://pypi.org/simple" }
 dependencies = [
    { name = "aiohttp" },
@@ -4333,9 +4333,9 @@ dependencies = [
    { name = "pydantic" },
    { name = "rich" },
 ]
-sdist = { url = "https://files.pythonhosted.org/packages/75/33/6091503673fb3c143293e9639bafe3acb6e7406e0ace1f893c34cbe22912/praisonaiagents-0.5.3.tar.gz", hash = "sha256:4163b00c28ef5e5d1a75411eb4b72fc7749790e52cdaa4bfc98c9a747fe42140", size = 780945, upload-time = "2025-12-31T16:36:45.335Z" }
+sdist = { url = "https://files.pythonhosted.org/packages/4b/3c/a25740d7e3da4ebea0d83319d412650a016d91b6eb2f2653df1b28a3882f/praisonaiagents-0.6.0.tar.gz", hash = "sha256:f83528f6f1c0adea17cea25a1ea93f660e46ba43002e354717501e82112d9a96", size = 780850, upload-time = "2025-12-31T18:50:35.35Z" }
 wheels = [
-    { url = "https://files.pythonhosted.org/packages/32/a2/459f111873f1751a34bc780e1b3a2cd264cf580f35acff868fa35aa9d917/praisonaiagents-0.5.3-py3-none-any.whl", hash = "sha256:35bd6c26dfff6c9e77575f72a2f077cdbcc0544d0d45b0ed29fb00ca96c3719c", size = 724879, upload-time = "2025-12-31T16:36:43.383Z" },
+    { url = "https://files.pythonhosted.org/packages/8e/18/f4f8360024ef26806831dde4ba3dad23950a4d94dde86085dde510ef1833/praisonaiagents-0.6.0-py3-none-any.whl", hash = "sha256:416c4550b6460b0866c1d59b284aacb86f23d849b38e75542f7056e530e84d80", size = 724880, upload-time = "2025-12-31T18:50:33.21Z" },
 ]

 [[package]]
Author	SHA1	Message	Date
Mervin Praison	3386d25022	Release v3.0.2	2026-01-02 04:33:19 +00:00
Mervin Praison	6dca2ace8a	feat(tui): queue editing + update_input method ## Queue Editing - Added update_input method to QueueScheduler for editing queued runs - Added update_input method to QueueManager - Added handler for QueuePanelWidget.EditRequested in TUIApp - Only queued (not running) items can be edited ## Queue-First Architecture Support - Queue items can be edited before processing - EditRequested message already existed in QueuePanelWidget 31 tests passing	2026-01-02 04:24:10 +00:00
Mervin Praison	63548739e8	feat(tui): session history persistence with SessionStore ## Session History Implementation - Created SessionStore class for in-memory chat history persistence - Added chat_history field to QueuedRun model - Modified queue worker to inject chat_history into Agent - Updated TUIApp to store user/assistant messages in SessionStore - Pass chat_history to queue when submitting messages ## How It Works 1. User sends message -> stored in SessionStore 2. Queue submission includes chat_history from SessionStore 3. Worker injects chat_history into Agent.chat_history 4. Agent has full conversation context 5. Response stored back in SessionStore ## Verified - Agent remembers previous messages in session - History format correct for LLM (role + content) - 31 tests passing	2026-01-02 04:22:26 +00:00
Mervin Praison	3e046537f4	fix(tui): sessions screen loading + update_sessions_screen method - Call _update_sessions_screen() after pushing sessions screen - Add _update_sessions_screen() method to load sessions from persistence - Fallback to showing current session if persistence fails 31 tests passing	2026-01-02 04:02:55 +00:00
Mervin Praison	4a34677567	fix(tui): compose() NoneType error + add missing slash commands ## Fixes 1. ChatWidget.compose() now yields empty generator instead of None 2. Added 7 missing slash commands: /tokens, /plan, /map, /undo, /diff, /commit, /tools ## Commands Added - /tokens - Show token usage breakdown - /plan - Show/create task plans - /map - Show repository map - /undo - Undo last change - /diff - Show diff of changes - /commit - Commit changes - /tools - Show available tools 31 tests passing	2026-01-02 04:01:25 +00:00
Mervin Praison	19e91cb669	fix(tui): proper scrollbar + non-blocking input ## Scrollbar Fix (VerticalScroll) - ChatWidget now inherits from VerticalScroll (not ScrollableContainer) - Messages mount directly to container (removed inner Vertical) - scroll_end() called after adding messages and during streaming - CSS has overflow-y: auto for proper scrollbar display ## Non-Blocking Input (@work decorator) - Message submission uses run_worker() for background execution - _process_message_submission decorated with @work - User can type while agent is processing - Multiple messages can be queued ## Root Causes Fixed 1. Inner Vertical container had overflow:hidden - removed it 2. Message handler was async blocking - now uses worker pattern Based on Textual documentation: - VerticalScroll: container with vertical layout and auto Y scrollbar - @work decorator: runs coroutine in background without blocking UI 31 tests passing	2026-01-02 03:56:56 +00:00
Mervin Praison	5a6125021a	fix(tui): add scrollbar to chat + fix session history ## Scrollbar Fix - ChatWidget now inherits from ScrollableContainer - Scrollbar visible in chat area (like queue panel) ## Session History Fix - Pass session_id from TUI app to queue submit - Pass session_id in agent_config - Worker now passes session_id to Agent constructor - Agent will now use session history for context Flow: TUI.session_id → queue.submit() → worker → Agent(session_id=...) 31 tests passing	2026-01-02 03:44:55 +00:00
Mervin Praison	da26887f24	feat(tui): NEWEST-AT-TOP rendering + fix @ typing blockage ## UX STRATEGY CHANGE (Stabilization) ### Newest-at-Top Rendering - Messages now render NEWEST at TOP - No scroll dependency - latest always visible - Streaming updates visible immediately - ChatWidget inherits from Vertical (not ScrollableContainer) ### @ Typing Not Blocked - @ handler no longer mounts focus-stealing popup - Users can freely type @path/to/file - Manual path entry works without interruption ### Verified Working - ✅ New messages appear at TOP - ✅ Latest response always visible - ✅ Streaming updates visible - ✅ Can type while agent runs - ✅ Queue increments immediately - ✅ Backslash popup works - ✅ @ does NOT block typing - ✅ 31 tests passing	2026-01-02 03:38:13 +00:00
Mervin Praison	1b2f7139f9	docs(tui): update help screen with auto-trigger features - Document @ auto-show file browser - Document / auto-show command discovery - Update backslash description to 'auto-show'	2026-01-02 03:25:17 +00:00
Mervin Praison	aafbc86180	feat(tui): auto-trigger popups for backslash, @, and / commands ## Auto-Trigger Features (Claude Code-style UX) ### Backslash (\) Auto-Popup - Typing \ immediately shows command popup - Clears backslash from input - No manual trigger needed ### @ File Autocomplete - Typing @ shows file/directory popup - Fuzzy search files in workspace - Supports ~/ and absolute paths - Reuses FileSearchService from at_mentions.py ### / Command Discovery - Typing / shows command popup with filter - Same unified command registry as backslash ## Queue Live Updates - Queue panel updates IMMEDIATELY when message queued - Updates on run completion and error - Reactive queued_count and running_count ## Composer Never Disabled - Input stays enabled during agent processing - Users can queue multiple messages ## New Files - widgets/file_popup.py: FilePopupWidget for @ autocomplete ## Tests - 31 tests passing (10 new tests) - TestAutoTriggerEvents: 4 tests - TestFilePopup: 4 tests - TestQueueLiveUpdates: 2 tests	2026-01-02 03:24:39 +00:00
Mervin Praison	3d4ca1b7c0	docs(tui): update help screen with safe keyboard defaults and backslash commands ## Help Screen Updates - Document backslash (\) command popup - Show safe keyboard defaults (no Ctrl/Fn by default) - Add command mode (: prefix) section - Add slash commands (/ prefix) section - Document optional Ctrl/Fn key env vars ## Keyboard Shortcuts Documented - \ - Show command popup - Enter - Send message - Shift+Enter - New line - q - Quit (when not typing) - ? - Help - : - Command mode - / - Search mode - Escape - Cancel/Close	2026-01-02 03:07:50 +00:00
Mervin Praison	1180aa38bd	feat(tui): add backslash command popup for Claude Code-style UX ## New Features ### Command Popup Widget (command_popup.py) - CommandPopupWidget: Overlay showing all available commands - Triggered by backslash (\) key - matches Claude Code UX - Searchable/filterable command list - Keyboard navigation (arrows, Enter, Escape) - 16 default commands with aliases ### MainScreen Integration - Backslash binding shows command popup - Escape dismisses popup or cancels input - Command selection executes command - Popup layer CSS for proper overlay positioning ## Commands Available - help, clear, model, cost, tokens, queue - cancel, settings, sessions, tools, exit - plan, undo, diff, commit, map ## Tests - 21 tests passing - TestCommandPopup: 4 new tests - TestMainScreenBackslash: 3 new tests ## Files Changed - NEW: widgets/command_popup.py - screens/main.py: Add popup integration - tests/unit/tui/test_tui_widgets.py: Add tests	2026-01-02 03:06:44 +00:00
Mervin Praison	65f65e5c7b	feat(tui): command mode routing and safe keyboard defaults ## Changes ### Command System (: and / prefix) - Both : and / now work as command prefixes - Local command handling in MainScreen._handle_local_command() - Commands: :quit, :clear, :help, :tools, :queue, :settings, :cancel - Aliases work: :q, :cl, :h, :t, :qu ### Keyboard Defaults - TUIApp bindings now hidden in footer (show=False) - Ctrl+Q/C kept for emergency exit but not displayed - MainScreen shows safe defaults: q, :, ?, Esc ### Help Text Updated - Shows safe keyboard shortcuts - Documents command mode usage - Removed references to Ctrl/Fn keys ## Files Changed - app.py: Hide Ctrl bindings in footer, update help text - screens/main.py: Add _handle_local_command() for : commands - widgets/composer.py: Support both / and : as command prefixes ## Tests - 14 tests passing - Command parsing verified - TUI app startup verified	2026-01-02 02:48:50 +00:00
Mervin Praison	f81a12d5e8	feat(tui): comprehensive UX fixes - safe keyboard defaults, auto-scroll, tools panel BREAKING CHANGE: Keyboard shortcuts changed to safe defaults ## Changes ### 1. Auto-scroll Fix - Use call_after_refresh() to ensure scroll happens after DOM updates - Fixed in ChatWidget.add_message() and update_streaming() ### 2. Tools Panel Fix - Added ScrollableContainer for tools list (no more truncation) - Shows count of available tools - All 13 tools now visible with scrolling ### 3. Safe Keyboard Defaults (NO Ctrl/Fn keys by default) - Single-char shortcuts: q (quit), : (command mode), ? (help) - Command mode: :quit, :clear, :help, :tools, :queue, :settings - Ctrl keys (^Q, ^C, ^L) - DISABLED by default (opt-in) - Function keys (F1-F5) - DISABLED by default (opt-in) ### 4. Configuration System - New TUIConfig class in config.py - Environment variables: - PRAISONAI_TUI_CTRL_KEYS=1 (enable Ctrl keys) - PRAISONAI_TUI_FN_KEYS=1 (enable Fn keys) - PRAISONAI_TUI_LEADER=; (custom leader key) ### 5. Non-blocking Input - Queue accepts messages while agents are running - Already implemented in previous commit ## Files Changed - widgets/chat.py - Auto-scroll fix - widgets/tool_panel.py - ScrollableContainer, show all tools - widgets/composer.py - set_text() method for command mode - screens/main.py - Safe default bindings, command mode actions - config.py - NEW: TUIConfig, command parsing, help text ## Tests - 14 tests passing - New tests for TUIConfig and safe bindings	2026-01-02 02:39:16 +00:00
Mervin Praison	cf6f5fbefe	feat(tui): fix help screen, Enter key, and add TDD tests Fixes: - Add HelpScreen for F1 key (was missing, caused KeyError) - Fix Enter key to send message using custom SubmitTextArea class - Fix Shift+Enter to insert newline properly - Register HelpScreen in TUIApp.SCREENS Tests: - Add TDD test suite for TUI widgets (12 tests) - Test Enter key submission - Test Shift+Enter newline - Test F1 help screen - Test Ctrl key bindings All 13 default tools (ACP, LSP, Basic) load correctly. Queue system tested and working with real API keys.	2026-01-02 02:23:02 +00:00
Mervin Praison	908d7710db	feat(tui): improve TUI usability and add tools panel display - Enter key now sends message (was Ctrl+Enter) - Shift+Enter inserts newline without backslash - Allow typing while response is generating (queue multiple messages) - Display available tools in Tools panel - Add edit/delete functionality to queue panel - Auto-scroll already working in chat widget TUI Architecture: - TUIApp -> QueueManager -> WorkerPool -> Agent (from praisonaiagents) - Tools stored in runtime registry to avoid JSON serialization issues	2026-01-02 02:08:22 +00:00
Mervin Praison	7d5b41e31f	fix: use 'llm' parameter instead of 'model' for Agent initialization in queue worker - Agent class uses 'llm' parameter, not 'model' - This fixes 'unexpected keyword argument model' error in TUI mode	2026-01-01 19:57:31 +00:00
Mervin Praison	e46a0bc9f3	Release v3.0.1	2025-12-31 19:29:02 +00:00
Mervin Praison	911483c92c	Release v3.0.0	2025-12-31 18:55:59 +00:00