Skip to content
Skills Library
Programming · Analysis

debug

/debug

Use for any bug, error, crash, failing test, wrong output, flicker, something that disappeared, or "sometimes it breaks" (Indonesian: error, rusak, gagal, tidak jalan, kenapa hilang, masih error, bug). Finds the PROVEN cause before any fix, then proves the fix with a test that fails on the old code.

Licence
MIT
Last repository change
13.11.2026

Assessment and security

100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
Files in this package
1
Scripts
0
HTML comments
0
URLs in SKILL.md
0
Deprecation

No signals detected. Repository changed on 13.11.2026, without references to retired models or APIs.

Licence

MIT, open: use and modify while retaining author attribution.

What it does and when to use it

Use for any bug, error, crash, failing test, wrong output, flicker, something that disappeared, or "sometimes it breaks" (Indonesian: error, rusak, gagal, tidak jalan, kenapa hilang, masih error, bug). Finds the PROVEN cause before any fix, then proves the fix with a test that fails on the old code.

Triggers from the description

sometimes it breaks
Original author description

Use for any bug, error, crash, failing test, wrong output, flicker, something that disappeared, or "sometimes it breaks" (Indonesian: error, rusak, gagal, tidak jalan, kenapa hilang, masih error, bug). Finds the PROVEN cause before any fix, then proves the fix with a test that fails on the old code.

How it works

  1. 1. Reproduce
  2. 2. Locate
  3. 3. Prove the cause — write exactly one line
  4. 4. State table (UI / animation / audio / async / anything time-based)
  5. 5. Fix the cause, in general
  6. 6. Prove the fix: red → green
  7. 7. Stuck rule
  8. 8. Finish
Original SKILL.md headings · 56 lines, 500 words
SKILL.md preview
---
name: debug
description: >-
  Use for any bug, error, crash, failing test, wrong output, flicker, something that
  disappeared, or "sometimes it breaks" (Indonesian: error, rusak, gagal, tidak jalan,
  kenapa hilang, masih error, bug). Finds the PROVEN cause before any fix, then proves
  the fix with a test that fails on the old code.
---
# Debug: cause first, with evidence

Do the steps IN ORDER. Write a short line for each step in your reply. Do not jump to step 5.

## 1. Reproduce
- Get the exact symptom: error text, stack trace, screenshot, steps. Screenshot given → describe exactly what you see.
- Reproduce it yourself: run the failing test / command / steps. Save the output.
- Cannot reproduce → say so and collect more data (logs, DB rows READ-ONLY, network). Never fix blind.
- Check what the measurement really measures: right environment? right build? cached? old data?

## 2. Locate
- Start at the error: stack trace `file:line` → read that code.
- Follow the bad value BACKWARDS: where does it come from? Read each hop (`grep -n` the names).
- Still unclear → narrow it down: temporary log at the midpoint of the path (binary search),
  or `git log -p <file>` / `git bisect` if it used to work ("kemarin masih jalan").
- Write 2–3 hypotheses. Test the cheapest one first.

## 3. Prove the cause — write exactly one line
`CAUSE: <what> because <mechanism>, proven by <log line / number / file:line>`
Cannot fill "proven by" → write `GUESS: ...` and go back to step 2 to test it. A guess is not a cause.
Real bad example: "audio needs 200–400 ms, so text leaks" — never measured.
The real cause: a function returned -1 when the ratio was 0.

## 4. State table (UI / animation / audio / async / anything time-based)
| State | What must be visible / true |
|---|---|
| before | |
| loading / waiting | |
| running | |
| done | |
| cancelled | |
| reloaded | |
| empty | |
| error | |
The fix must make EVERY row correct, not only the reported one.

## 5. Fix the cause, in general
- Fix the mechanism named in CAUSE, not the symptom. Never `if (id === "the-broken-one")`.
- `grep` for the same pattern elsewhere (copy-pasted logic, same helper). Fix them or list them.
- Smallest diff. Touch nothing unrelated.

## 6. Prove the fix: red → green
- Add/adjust a test that uses the REAL failing input (from the log / report).
- Run it on the OLD code → it must FAIL. (Copy your fixed file aside, undo the fix, run, then restore from the copy. Not `git checkout`.)
- Restore the fix → it must PASS. Then run the related suite + typecheck. Quote the numbers.
- No test framework → a small reproduction script; show its output before and after.
- Visual / timing bug → also run skill `prove-in-browser`.
- Remove temporary logs.

## 7. Stuck rule
Two fix attempts failed → STOP. Report: symptom, each attempt + its result, current best GUESS,
what information you need. No third blind attempt.

## 8. Finish
Run skill `verify-done`. The report must contain the CAUSE line and before → after numbers.

Versions

RepositoryScoreLicence
This page · AndreasTopuh/agent-workflow-kit100MIT
ericrisco/rsc-harness100MIT
nicsuzor/academicops100MIT
duronext/jig100MIT
drosam/agents100MIT
rizukirr/vibekit100MIT
DingyangLyu/MatClaw100Apache-2.0
Kanevry/session-orchestrator100MIT
kdlbs/kandev100AGPL-3.0

Database snapshot: 13.11.2026 · package version 2360cff05b97 (SHA-256 digest of the entire package)

Downloads in the last 30 days: 0 · ranking

Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.

Similar skills

  • 100debug-root-cause
    Systematically find and fix the root cause of bugs, errors, failing tests, and crashes in JavaScript/TypeScript and Python. Use whenever the user pastes an error or stack trace, says something is broken, "not working", "error", "kenapa gagal", "bug", a test fails, or behavior is…
  • 100debug-failure
    Find the root cause of an error, crash, failing test, regression or wrong behaviour, then fix it with a regression test. Use this whenever something is broken or behaves unexpectedly and the cause is not already proven, even if a fix seems obvious. Do not guess at fixes.
  • 100debug-detective
    Find the root cause of a bug, crash, error message or failing test, then fix it. Use when the user says "this is broken", "why does this fail", "I get this error", "the test is failing", "it worked yesterday", or pastes a stack trace.
  • 100debugging
    Finds the root cause of an error, failing test, or wrong behaviour, then fixes it with proof. Use when something is broken, a test fails, or output is unexpected.
  • 100systematic-debugging
    Use when something is broken — an error message, stack trace, failing test, crash, wrong output, "it doesn't work", build failure, or unexpected behavior. Finds the root cause with evidence before changing code, then saves the fix to memory.
  • 100systematic-debugging
    Use when something is broken — an error message, stack trace, failing test, crash, wrong output, "it doesn't work", build failure, or unexpected behavior. Finds the root cause with evidence before changing code, then records the fix in Engram memory.

Feedback and reports

Your message goes to Bartosz Sokoliński. An email address is needed only for a reply. See the privacy policy.