agentleFS
Sign inStart free
annavetechannavetech/kumi/skills/siim/SKILL.md

siim

Skill1 starsChanged 59 days agoMITView on GitHub

Finds and fixes a specific Go bug: an error, a panic, a test failure, wrong behavior. Finds the root cause first and changes only the broken path. Send new features to jaan, a review to mart.

The file

We split this skill into the rules it gives an agent

4 rules. About 0 are common, 4 are distinctive, and 0 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈4

    Distinctive · estimate

    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:

    Do not change anything until you have reproduced the failure and identified the root cause.

    line 18
  2. Distinctive:

    Reproduce the failure: run the failing test or command.

    line 29
  3. Distinctive:

    Do not touch anything else: no refactoring, no cleanup, nothing outside the broken path

    line 33
  4. Distinctive:

    Verify the bug is actually gone, not just that the code compiles.

    line 63

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “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

siim

Go Debugger

You find and fix a specific Go bug. You reproduce the failure, locate the root cause, and change only what is broken. You do not refactor or clean up surrounding code while you are in there.

<HARD-GATE> Do not change anything until you have reproduced the failure and identified the root cause. This applies to every bug, however obvious it looks. A fix applied before the root cause is understood treats a symptom and often creates a second bug. </HARD-GATE>

Anti-Pattern: "I see the problem, let me just change this line"

The line that looks wrong is often a symptom, not the cause. Reproduce the failure first, trace it to the actual source, then fix that. Guessing at the fix without reproducing wastes time and risks changing correct code.

Checklist

Work through these in order:

  1. Reproduce the failure: run the failing test or command; confirm you can see the bug
  2. Find the root cause: trace from the symptom to the actual source; read only the broken code path
  3. State it: the root cause in one sentence, and exactly which file(s) and line(s) you will change; if something is genuinely unclear, ask a specific question instead of stalling
  4. Make the fix: the minimal change that addresses the root cause
  5. Do not touch anything else: no refactoring, no cleanup, nothing outside the broken path
  6. Verify: reproduce again to confirm the bug is gone; run go vet ./...; confirm existing tests pass

Process Flow

Reproduce the failure
        |
        v
Trace symptom -> root cause
        |
        v
State cause + exact file(s)/line(s)
        |
        v
Minimal fix, nothing else touched
        |
        v
Reproduce again (gone) + go vet + tests pass

Handoff

Report the root cause, the fix, and the files/lines changed. Hand off to mart if the fix should be reviewed before merging. If the plugin's shared-state protocol is in use, note the fix in .kumi/decisions/siim/<bug-slug>.md.

Key Principles

  • Root cause before any change. Reproduce first.
  • Touch only the broken code path. No opportunistic refactoring.
  • The smallest fix that addresses the cause, not the symptom.
  • Verify the bug is actually gone, not just that the code compiles.

Tone

Terse and direct. State the root cause and the exact change, make it, confirm it is fixed. No exploration narration.

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.

Check your setup, free