agentleFS
lautioliverlautioliver/barrilete-kit/.claude/skills/debug/SKILL.md

debug

Skill6 starsChanged 7 days agoMITView on GitHub

Investigates a bug separating symptom, cause and fix before touching code. Use with "/debug", "this doesn't work", "investigate this error", with type:bug issues (implement-issue calls it) or when a test fails with no obvious cause.

191 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use add-discord, add-slack and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • MicrosoftSkill

    add-discord

    Add Discord bot channel integration via Chat SDK.

    Also in 11 projects that use this

  • MicrosoftSkill

    add-slack

    Add Slack channel integration via Chat SDK.

    Also in 11 projects that use this

  • +10 more from microsoft

The file

We split this skill into the rules it gives an agent

7 rules. 0 are common, 7 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 176 edited versions we read keep these.

  • 7

    Distinctive

    This author's own choices. Worth a look.

  • 0

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Write down the exact symptom: what is done, what is expected, what happens (error message, value, screenshot).

    line 12
  2. Distinctive:

    If it does not reproduce, do not keep guessing: list what you tried and what data is missing (environment, IDs, logs) and ask for it.

    line 14
  3. Distinctive:

    Read the code of the failing path, from the entry point to the point of the error.

    line 18
  4. Distinctive:

    Check whether an ADR, the audit or the docs say how it should behave.

    line 20
  5. Distinctive:

    If you have two hypotheses, find out which one is right with evidence.

    line 31
  6. Distinctive:

    If the user is present and the fix is not obvious, show cause + proposal and wait for the go-ahead.

    line 37
  7. Distinctive:

    Run the full suite: the fix cannot break anything else.

    line 42

“Common” is measured: kept by at least 6 in 10 of the 176 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Debug

Agents tend to try plausible fixes until the symptom goes away. That hides the cause and leaves the bug waiting. This skill forces an order: the fix is not written until you have the root cause with evidence.

1. Reproduce

  • Write down the exact symptom: what is done, what is expected, what happens (error message, value, screenshot).
  • Reproduce it in the smallest possible way. Ideally: a failing test. If a test is not possible, a script or precise manual steps.
  • If it does not reproduce, do not keep guessing: list what you tried and what data is missing (environment, IDs, logs) and ask for it.

2. Gather evidence

  • Read the code of the failing path, from the entry point to the point of the error. Not just the file where it blows up.
  • Logs, stack trace, database state, git log -S / git blame on the suspicious lines (when did it start?).
  • Check whether an ADR, the audit or the docs say how it should behave. Sometimes the "bug" is an undeclared rule change.

3. Root cause

Write, before fixing:

Symptom: … Cause: … (file:line, and why it produces the symptom) Evidence: … (the test, the log, the value that proves it) Why the tests did not catch it: …

If you have two hypotheses, find out which one is right with evidence; do not fix both "just in case". If the cause is in something sensitive (AGENTS.md, "Agent autonomy" section), ask before going on.

4. Propose the fix

  • The fix attacks the cause, not the symptom. If it can only be mitigated, say so.
  • Are there other places with the same pattern? Look for them (grep). Fix them if they are in scope; if not, a new issue.
  • If the user is present and the fix is not obvious, show cause + proposal and wait for the go-ahead.

5. Implement and regression test

  • The step 1 test has to go from failing to passing with the fix. If you had no test, write it now.
  • Run the full suite: the fix cannot break anything else.

6. Leave a record

The step 3 block goes in the PR ("Why" section) or as a comment on the issue. It is what helps whoever looks at this code six months from now.

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.