Skip to content
OpenAgentsbeta

Log Triage

The failure mode of reading logs is anchoring on the first stack trace and explaining everything else in its terms. This skill works the other way: establish the timeline first, separate what changed from what is always noisy, and only then form hypotheses, ranked by evidence and each with a cheap test attached.

When to use

  • An incident where you have logs and no diagnosis.
  • A failure that reproduces intermittently and you need to find the pattern.
  • Any time you catch yourself already sure of the cause after reading one line.

Install

bash
npx openagents-cli add openagents/log-triage
RuntimeInstalled to
claude-code.claude/skills/log-triage/
cursor.cursor/rules/log-triage/
codex.codex/skills/log-triage/
openai-agentsagents/log-triage/
langgraphgraphs/log-triage/
generic.openagents/log-triage/

What is in the package

  • SKILL.md - the triage procedure, and the mistakes it exists to prevent.
  • templates/triage.md - the output format: timeline, hypotheses, tests.

License

MIT

Keyboard shortcuts

Focus search
/
Go to Explore
ge
Go to Home
gh
Go to Tags
gt
Go to Collections
gc
Show this help
?
Close suggestions or this dialog
Esc