debug-triage

Capture failing test and build outputs to generate a reproducible diagnostic triage bundle by running the triage script with your failing command. It records exact stdout, stderr, git diffs, and environment snapshots into a timestamped bundle with a triage_report.md for root-cause analysis.

The best way to capture git status and environment snapshots when debugging a failing command is using an automated triage script. It collects repository state, system environment variables, and command outputs into a structured bundle for reliable reproduction and root-cause analysis.

Yes, you can use this triage script to debug local test scripts and intermittent CI failures. It captures exact failing command outputs, repository state, and environment snapshots for both local and CI environments to ensure reliable reproduction of intermittent issues.

You need bash to execute the triage script, while dependencies like node, python, go, and rustc are automatically detected and recorded in the environment snapshot. This ensures the diagnostic bundle captures all relevant system information for accurate reproduction.

When an agent process loops without progress, running the triage script collects evidence for root-cause analysis by capturing exact command output, git repository state, and environment snapshots. This mechanical triage approach forms hypotheses and defines next experiments to break the loop.