Home

Agents Discussion

Multi-agent automated root cause analysis for production incidents.

Agents Discussion runs a structured debate between three specialized LLM agents — a diagnostician, a skeptic, and a moderator — to diagnose technical issues through iterative hypothesis generation, critique, and moderation, grounded in real read-only tools (SSH, kubectl, Prometheus, Loki, SQL EXPLAIN, Git) with human-in-the-loop oversight.

Documentation

• Architecture — design, data flow, state schema, graph topology
• Usage — CLI, web UI, templates, tools, examples
• Operations — deployment, configuration, monitoring, troubleshooting
• Coding style — conventions, patterns, versioning & releases

Reference (generated from the code)

• REST API
• CLI
• Configuration
• Diagnostic tools
• SSE event types

The Reference section is generated by scripts/gen_docs.py and verified in CI, so it cannot drift from the application.

Source code: github.com/luisgf/ia_agents_discussion