6 Commits

Author SHA1 Message Date
John Doe abba890940 P0-1: Implement Heretek Observability Layer (Langfuse Integration)
Features:
- Langfuse client with triad consciousness metrics (GWT, IIT, AST)
- Trace context propagation for A2A messages
- Custom metrics exporter for deliberation, consensus, consciousness
- Dashboard sync for real-time Control Dashboard updates
- Gateway instrumentation for WebSocket RPC and A2A tracing
- LiteLLM integration for token/cost/latency tracking

Components:
- modules/observability/langfuse-client.js - Triad-aware Langfuse wrapper
- modules/observability/trace-context.js - A2A message propagation
- modules/observability/metrics-exporter.js - Custom metrics collection
- modules/observability/dashboard-sync.js - Dashboard integration
- modules/observability/gateway-instrumentation.js - Gateway tracing
- modules/observability/litellm-integration.js - LiteLLM observability
- modules/observability/index.js - Core module exports
- docs/observability/ARCHITECTURE.md - Architecture documentation
- docs/observability/SETUP.md - Installation guide
- docs/observability/METRICS.md - Metrics reference
- docs/observability/TRACING.md - Distributed tracing guide
- observability.json - Configuration file

Supports offline mode (local Langfuse) and cloud mode (Langfuse Cloud).
Heretek-native implementation with triad deliberation tracing.
2026-04-01 14:05:02 -04:00
John Doe 3ab39ccf91 Phase 4: Add comprehensive documentation and operations runbooks
- Add HERETEK_README.md: Comprehensive guide with architecture, configuration, and troubleshooting
- Add RUNBOOK_A2A_DEBUG.md: A2A protocol debugging procedures
- Add RUNBOOK_AGENT_OFFLINE.md: Agent offline detection and recovery
- Add RUNBOOK_GATEWAY_FAILURE.md: Gateway failure recovery procedures
- Add RUNBOOK_APPROVAL_ISSUES.md: Liberation plugin and approval bypass troubleshooting
- Add DEPLOYMENT_HERETEK.md: Docker Compose, Kubernetes, and bare metal deployment guides
- Add MIGRATION_FROM_UPSTREAM.md: Upstream to Heretek migration guide
- Update docs/operations/README.md: Add Heretek-specific runbooks and quick reference table

All runbooks include:
- Clear step-by-step procedures
- CLI commands with expected output
- Escalation paths
- References to source files
2026-04-01 13:24:54 -04:00
John Doe 8deeddb3a3 feat(monitoring): Add Matrix keepalive daemon, metrics server, and routing guide
- scripts/matrix-keepalive.sh: periodic keepalive for Matrix connections
- scripts/metrics-server.py: lightweight HTTP metrics endpoint
- docs/operations/MATRIX-ROUTING-GUIDE.md: routing documentation
- scripts/export-agent-metrics.sh: streamlined (legacy cleanup)
2026-04-01 10:33:27 -04:00
John Doe a9aa017b5f docs(operations): Document Matrix communication setup
- Records existing Tabula_Myriad Matrix rooms (4 rooms)
- Plans 4 new coordination rooms (triad-private, operations, monitoring, general)
- Documents agent Matrix IDs (@agent:node.heretek.one)
- Specifies configuration requirements for each agent

Server: node.heretek.one
2026-04-01 02:23:28 -04:00
John Doe d78099381a docs(operations): Add multi-session mode documentation
- Documents 8 active agent sessions for Control-UI visibility
- Lists all 23 agents (8 multi-session + 15 single-session)
- Explains session state file structure
- Provides monitoring commands
- Notes A2A communication methods
- Outlines future enhancements (Matrix, WebSocket dashboard)

Agents with visible sessions:
- alpha, beta, charlie (Triad)
- coder, examiner, explorer, oracle, sentinel (Operational)
2026-04-01 01:55:30 -04:00
John Doe a43521056a docs(operations): Add agent monitoring guide
- Documents Langfuse integration and limitations
- OpenClaw gateway health checks
- Plugin monitoring APIs (consciousness, liberation, hybrid-search)
- System-level monitoring (Docker, resources)
- Troubleshooting procedures
2026-04-01 01:49:00 -04:00