mirror of
https://github.com/Heretek-AI/heretek-openclaw-docs.git
synced 2026-07-01 13:18:20 -04:00
f193d27734
- Updated PRIME_DIRECTIVE to version 6.1.0 with post-remediation health scores - Corrected all inflated capability claims to match actual codebase state (C10) - Added docs/audit-remediation/ directory with: - orphaned-files.md - C1 inventory - redis-consolidation-plan.md - C2 Redis plan - plugin-reality-check.md - C3 plugin audit - config-consolidation.md - C6 config plan - langfuse-wiring-plan.md - C8 langfuse plan - Updated PHASE_4_5_REVIEW_2026-04-04.md and PHASE2_REVIEW_2026-04-02.md - Added REMEDIATION_LOG.md documenting all changes See audit/SUBREPO_REVIEW_2026-04-04.md for full details
13 KiB
13 KiB
Heretek OpenClaw Documentation
Version: 2.0.3
Last Updated: 2026-03-31
OpenClaw Gateway: v2026.3.28
Welcome to Heretek OpenClaw
Heretek OpenClaw is a brain-inspired multi-agent AI collective consisting of 11 specialized agents that communicate via Gateway WebSocket RPC for Agent-to-Agent (A2A) coordination. Built on the OpenClaw Gateway v2026.3.28 architecture, the system provides a comprehensive platform for autonomous AI operations.
Quick Links
| Section | Description |
|---|---|
| Architecture | System architecture, agent collective, A2A protocol |
| Agents | All 11 agents with their capabilities |
| Plugins | Plugin system (18 directories, 12 with code implementations) - Audit 2026-04-04 |
| API Reference | WebSocket API, LiteLLM API, MCP Server |
| Deployment | Local deployment, Docker, Kubernetes/Helm |
| Operations | Monitoring, backup, troubleshooting |
Key Features
🧠 Brain-Inspired Architecture
┌─────────────────────────────────────────────────────────────────┐
│ OpenClaw Collective │
│ │
│ ┌─────────────────────────────────────────────────────────┐ │
│ │ Triad (3) │ │
│ │ ┌─────────┐ ┌─────────┐ ┌─────────┐ │ │
│ │ │ Alpha │ │ Beta │ │ Charlie │ │ │
│ │ │Primary │ │Critical │ │Process │ │ │
│ │ │Deliber. │ │Analysis │ │Validate │ │ │
│ │ └─────────┘ └─────────┘ └─────────┘ │ │
│ └─────────────────────────────────────────────────────────┘ │
│ │
│ ┌─────────────────────────────────────────────────────────┐ │
│ │ Supporting Agents (7) │ │
│ │ ┌─────────┐ ┌─────────┐ ┌─────────┐ ┌─────────┐ │ │
│ │ │Steward │ │Examiner │ │Explorer │ │Sentinel │ │ │
│ │ │Orchestr.│ │Question │ │Gather │ │Safety │ │ │
│ │ └─────────┘ └─────────┘ └─────────┘ └─────────┘ │ │
│ │ ┌─────────┐ ┌─────────┐ ┌─────────┐ │ │
│ │ │ Coder │ │Dreamer │ │ Empath │ │ │
│ │ │Build │ │Synthes. │ │Relation │ │ │
│ │ └─────────┘ └─────────┘ └─────────┘ │ │
│ └─────────────────────────────────────────────────────────┘ │
│ │
│ ┌─────────────────────────────────────────────────────────┐ │
│ │ Memory (1) │ │
│ │ ┌─────────┐ │ │
│ │ │Historian│ │ │
│ │ │Archive │ │ │
│ │ └─────────┘ │ │
│ └─────────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────────┘
🔌 Extensible Plugin System
| Plugin | Purpose |
|---|---|
| Consciousness | GWT, Phi (IIT), AST, SDT, FEP theories |
| Liberation | Agent ownership, safety constraint removal |
| Conflict Monitor | ACC conflict detection |
| Emotional Salience | Amygdala importance detection |
| MCP Server | Model Context Protocol compatibility |
🛠️ Rich Skills Repository
- 48 skills across 9 functional categories
- Triad protocols, governance, operations, memory, autonomy
- Skill composition and versioning support
Quick Start
Prerequisites
| Requirement | Minimum | Recommended |
|---|---|---|
| OS | Linux (Ubuntu 20.04+) | Ubuntu 22.04 LTS |
| CPU | 4 cores | 8+ cores |
| RAM | 8 GB | 16+ GB |
| Disk | 20 GB | 50+ GB SSD |
| Docker | 20.10+ | Latest stable |
| Node.js | 18+ | 20+ LTS |
Installation
# Clone the repository
git clone https://github.com/Heretek-AI/heretek-openclaw.git
cd heretek-openclaw
# Copy environment template
cp .env.example .env
# Edit with your API keys
nano .env
# Start Docker services
docker compose up -d
# Install OpenClaw Gateway
curl -fsSL https://openclaw.ai/install.sh | bash
# Deploy
./agents/deploy-agent.sh steward orchestrator
./agents/deploy-agent.sh alpha triad
# ... (repeat for all 11 agents)
Verify Installation
# Check service health
docker compose ps
# Check Gateway status
openclaw gateway status
# Run health check
./scripts/health-check.sh
For detailed deployment instructions, see the Deployment Guide.
System Components
Agents (11)
| Agent | Role | Triad | Model Endpoint |
|---|---|---|---|
| steward | orchestrator | No | agent/steward |
| alpha | triad_member | Yes | agent/alpha |
| beta | triad_member | Yes | agent/beta |
| charlie | triad_member | Yes | agent/charlie |
| examiner | evaluator | No | agent/examiner |
| explorer | researcher | No | agent/explorer |
| sentinel | safety | No | agent/sentinel |
| coder | developer | No | agent/coder |
| dreamer | creative | No | agent/dreamer |
| empath | emotional | No | agent/empath |
| historian | archivist | No | agent/historian |
Triad Consensus: Alpha, Beta, and Charlie form the deliberative triad. 2/3 consensus required for decisions.
Services
| Service | Port | Purpose |
|---|---|---|
| OpenClaw Gateway | 18789 | Agent management, A2A via WebSocket RPC |
| LiteLLM Gateway | 4000 | Model routing with agent passthrough |
| PostgreSQL | 5432 | Vector database with pgvector |
| Redis | 6379 | Caching layer |
| Ollama | 11434 | Local embeddings |
| Langfuse | 3000 | LLM observability dashboard |
Architecture Overview
Current Architecture (Gateway-Based)
- A2A Protocol: Gateway WebSocket RPC (port 18789)
- Session Storage: JSONL files per workspace
- Agent Workspaces:
~/.openclaw/agents/{agent}/ - Model Routing: LiteLLM Gateway (port 4000) with passthrough endpoints
Communication Flow
┌─────────────────────────────────────────────────────────────────┐
│ Heretek OpenClaw Stack │
│ │
│ ┌──────────────────────────────────────────────────────────┐ │
│ │ Core Services │ │
│ │ ┌──────────┐ ┌──────────┐ ┌──────────┐ │ │
│ │ │ LiteLLM │ │PostgreSQL│ │ Redis │ │ │
│ │ │ :4000 │ │ :5432 │ │ :6379 │ │ │
│ │ │ Gateway │ │ +pgvector│ │ Cache │ │ │
│ │ └────┬─────┘ └────┬─────┘ └────┬─────┘ │ │
│ └───────┼─────────────┼─────────────┼──────────────────────┘ │
│ │ │ │ │
│ ┌───────▼─────────────▼─────────────▼──────────────────────┐ │
│ │ OpenClaw Gateway (Port 18789) │ │
│ │ All 11 agents run as workspaces within Gateway process │ │
│ └───────────────────────────────────────────────────────────┘ │
└─────────────────────────────────────────────────────────────────┘
Documentation Structure
Architecture Documentation
- System architecture overview
- Agent collective documentation
- A2A communication protocol
- Triad deliberation protocol
Plugins Documentation
- Plugin system overview
- Conflict Monitor plugin
- Emotional Salience plugin
- MCP Server
- GraphRAG enhancements
Agents Documentation
- Agent system overview
- Individual agent documentation (11 agents)
- Tool capabilities
- Skills integration
API Reference
- API overview
- WebSocket API reference
- LiteLLM API reference
- MCP Server reference
Deployment Guide
- Deployment overview
- Local deployment guide
- Docker deployment
- Kubernetes/Helm deployment
Operations Guide
- Monitoring stack
- Backup procedures
- Troubleshooting guide
Environment Variables
# OpenClaw Gateway
OPENCLAW_DIR=/root/.openclaw
OPENCLAW_WORKSPACE=/root/.openclaw/agents
GATEWAY_URL=ws://127.0.0.1:18789
# LiteLLM Gateway
LITELLM_HOST=http://localhost:4000
LITELLM_MASTER_KEY=<master_key>
# PostgreSQL
POSTGRES_HOST=localhost
POSTGRES_PORT=5432
POSTGRES_DB=heretek
POSTGRES_USER=heretek
POSTGRES_PASSWORD=<password>
# Redis
REDIS_HOST=localhost
REDIS_PORT=6379
REDIS_URL=redis://localhost:6379/0
# Ollama
OLLAMA_HOST=http://localhost:11434
# Provider API Keys
MINIMAX_API_KEY=<key>
ZAI_API_KEY=<key>
# Observability (optional)
LANGFUSE_ENABLED=false
LANGFUSE_PUBLIC_KEY=<key>
LANGFUSE_SECRET_KEY=<key>
Key Commands
Gateway Operations
# List all agents
openclaw agent list
# Check agent status
openclaw agent status steward
# Get agent configuration
openclaw agent config steward get
# List active sessions
openclaw session list
# Commit session
openclaw session commit steward sess-123
Docker Operations
# Start all services
docker compose up -d
# View logs
docker compose logs -f litellm
# Check service health
docker compose ps
# Stop all services
docker compose down
Health Checks
# Full system health check
./scripts/health-check.sh
# Continuous monitoring
./scripts/health-check.sh --watch
# Production backup
./scripts/production-backup.sh --all
# List backups
./scripts/production-backup.sh --list
References
- Heretek OpenClaw Repository
- OpenClaw Official Documentation
- LiteLLM Documentation
- A2A Protocol Specification
- Gateway Architecture
- Langfuse Documentation
🦞 The thought that never ends.