agentleFS
NemesLaszloNemesLaszlo/claude-code-harness/.claude/skills/fix-issues/SKILL.md

fix-issues

Skill1 starsChanged 12 days agoMITView on GitHub

Fix all issues from a cross-validation review summary, a /review verdict, or a Greptile review. Use when the user says fix issues, fix these, fix the review findings, address the issues, or after /implement produces a cross-validation summary.

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

11 rules. 0 are common, 9 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 3 edited versions we read keep these.

  • 9

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    If no arguments provided, find the most recent cross-validation summary, /review output, or /security-scan result in this conversation and …

    line 20
  2. Distinctive:

    Read the file and understand the surrounding code

    line 38
  3. Distinctive:

    Verify the fix doesn't break adjacent code

    line 40
  4. Distinctive:

    Keep changes minimal and focused on the reported issue

    line 53
  5. Distinctive:

    Don't refactor surrounding code unless it's part of the fix

    line 54
  6. Distinctive:

    Follow existing patterns in the file

    line 55
  7. Distinctive:

    If a fix requires changes in multiple files, make all related changes together

    line 56
  8. Very specific:

    If a fix extracts logic shared by two or more callers, follow /code-structure: one extraction, one caller replaced, verified, then the rest

    line 57
  9. Distinctive:

    After all fixes are applied:

    line 61
  10. Distinctive:

    Run tests if they exist

    line 62
  11. Distinctive:

    Do a quick scan of all modified files to ensure no regressions

    line 63

“Common” is measured: kept by at least 6 in 10 of the 3 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

Context

  • Current branch: !git branch --show-current || true
  • Modified files: !git status --short || true
  • Staged changes: !git diff --cached --stat || true

Your Task

Fix all issues identified in the most recent cross-validation summary or code review.

Specific issues to fix: $ARGUMENTS

If no arguments provided, find the most recent cross-validation summary, /review output, or /security-scan result in this conversation and address all Must Fix and Should Fix items.


Fix Process

Step 1: Parse the issues

Extract all issues from the review summary. Categorize by file and type:

  • Security issues (fix first)
  • Bug/correctness issues
  • Error handling gaps
  • Code quality issues
  • Performance issues

Step 2: Fix in order

For each issue:

  1. Read the file and understand the surrounding code
  2. Apply the fix following existing codebase patterns
  3. Verify the fix doesn't break adjacent code
  4. Move to the next issue

Agent Assignment:

  • Backend file issues → use Backend Architect agent
  • Frontend file issues → use Frontend Developer agent
  • Cross-cutting or refactoring → use Clean Code Architect agent
  • Security-specific fixes → ask the Security Vulnerability Scanner agent for the fix (it cannot edit files), then hand that fix to the Backend Architect or Frontend Developer that owns the file

Launch each agent through the Agent tool with its subagent_type (backend-architect, frontend-developer, clean-code-architect; security-vulnerability-scanner for guidance only) and give it the issue list for its files. Group issues by file so one agent handles one file. Do not apply fixes from the main context.

Fix Rules:

  • Fix the actual issue — don't just suppress warnings
  • Keep changes minimal and focused on the reported issue
  • Don't refactor surrounding code unless it's part of the fix
  • Follow existing patterns in the file
  • If a fix requires changes in multiple files, make all related changes together
  • If a fix extracts logic shared by two or more callers, follow /code-structure: one extraction, one caller replaced, verified, then the rest

Step 3: Verify

After all fixes are applied:

  1. Run tests if they exist
  2. Do a quick scan of all modified files to ensure no regressions
  3. Re-capture evidence for any behavior the fixes changed (/evidence-driven-testing)

Step 4: Report

## Fix Summary

### Issues Fixed
- [X/Y] Must Fix items resolved
- [X/Y] Should Fix items resolved
- [X/Y] Consider items addressed (if any)

### Changes Made
- `file:line` — [what was fixed]

### Issues Not Fixed (if any)
- [Issue] — [reason: needs human decision / out of scope / blocked by X]

### Verification
- Tests: [passed/failed/no tests]
- Regression check: [clean/issues found]
- Evidence: [path under .artifacts/ or "unchanged"]

### Next Step
`/evidence-driven-testing` if behavior changed, then `/ship`.

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.