agentleFS
Sign inStart free
cellrishi-codecellrishi-code/claude-skills-atlas/skills/coding/debug-systematically/SKILL.md

debug-systematically

Skill6 starsChanged 39 days agoMITView on GitHub

Diagnose software failures systematically instead of guessing. Use for bugs, failing tests, unexpected output, crashes, regressions, and inconsistent behavior.

5 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

6 rules. 0 are common, 6 are distinctive, and 0 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 6

    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:

    Treat debugging as evidence gathering and hypothesis elimination.

    line 8
  2. Distinctive:

    Test the cheapest hypothesis that can distinguish them.

    line 17
  3. Distinctive:

    Never change multiple unrelated things before verifying the cause.

    line 24
  4. Distinctive:

    Do not replace an error with silent fallback behavior unless intended.

    line 25
  5. Distinctive:

    Treat intermittent failures as evidence about timing, state, concurrency, caching, or environment.

    line 27
  6. Distinctive:

    If the cause cannot be established, label it unresolved rather than guessing.

    line 28

“Common” is measured: kept by at least 6 in 10 of the 5 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 Systematically

Treat debugging as evidence gathering and hypothesis elimination.

Process

  1. Reproduce the failure when possible.
  2. Capture the exact error, input, environment, and expected behavior.
  3. Minimize the failing case.
  4. Trace backward from the observable failure to the earliest incorrect state.
  5. Generate a small set of competing hypotheses.
  6. Test the cheapest hypothesis that can distinguish them.
  7. Apply the smallest correct fix.
  8. Re-run the original reproduction and relevant regression tests.
  9. Explain the root cause, not only the patch.

Rules

  • Never change multiple unrelated things before verifying the cause.
  • Do not replace an error with silent fallback behavior unless intended.
  • Preserve useful diagnostic information.
  • Treat intermittent failures as evidence about timing, state, concurrency, caching, or environment.
  • If the cause cannot be established, label it unresolved rather than guessing.

Output

Return:

  • Reproduction
  • Observed vs expected behavior
  • Root cause
  • Minimal fix
  • Verification
  • Remaining uncertainty

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.

Reviews can't be read right now.

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.

Check your setup, free