| name | rounds-help |
| description | Print available rounds skills, project directory layout, and a brief overview of what rounds does |
| user_invocable | true |
| args | [] |
Rounds: Help
Print a quick-reference guide for working with rounds inside the container.
Implementation
Output the following text exactly, with no additional commentary:
ROUNDS — Autonomous Error Diagnosis Agent
==========================================
Rounds is a lightweight daemon that watches your OpenTelemetry data, fingerprints
failure patterns, and uses LLM-powered analysis to diagnose root causes in your
codebase. It makes periodic passes over your telemetry — polling for errors,
tracing them through your system, recognizing recurring patterns, and building up
a persistent knowledge base of failure signatures and diagnoses. Think of it as a
nurse making rounds on your running software.
SKILLS AVAILABLE IN THE ROUNDS CONTEXT
---------------------------------------
/rounds-help — show this message
/rounds-investigate TRACE_ID — investigate a distributed trace
/rounds-list [STATUS] — list error signatures
/rounds-details SIGNATURE_ID — full signature details
/rounds-reinvestigate SIG_ID — re-run LLM diagnosis
/rounds-mute SIG_ID [REASON] — mute a signature
/rounds-resolve SIG_ID — mark a signature resolved
/rounds-daemon — start the polling daemon
/rounds-test — run pytest
/rounds-check — mypy + ruff
/rounds-explore [QUERY] — query telemetry: search logs, spans, traces
/rounds-assess [QUERY] — assess transaction types: instrumentation, code quality, usage, skill gaps
DIRECTORIES (inside container)
-------------------------------
/workspace/rounds — rounds source code (this project)
/workspace/target — target codebase being monitored and diagnosed