Autonomous Multi-Agent Incident Response and Reliability Orchestration Platform
Current Limitations: SentinelOps is currently a development prototype. Leading metric Root Cause Match is at 11.93% (lexical) / 14.68% (semantic) accuracy under simulation. Autonomy is disabled; human supervision is required for all actions.
cp .env.example .env
make up
curl http://localhost:8000/health
.venv/bin/pytestSentinelOps uses a cooperative multi-agent architecture powered by LangGraph. Details are documented in docs/architecture/current.md.
| Metric | Target | Current Value | Status |
|---|---|---|---|
| Root Cause Match (top-1) | >75% | 11.93% (lexical) / 14.68% (semantic) | Warning |
| Classification Accuracy | >90% | 99.17% | Passing |
| Evidence Grounding Score | >0.95 | 98.4% | Passing |
See docs/getting-started.md for full onboarding.