devattom/ci-fixer
skills/workflow-ci-fixer/SKILL.md
Automated CI/CD pipeline fixer - watches CI, fixes errors locally, commits, and loops until green. Use when CI is failing and you want to automatically fix and verify changes.
testing
Updated Apr 4, 2026
$ install --global
skillsauthnpx skillsauth add devattom/.claude ci-fixerInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: Apr 4, 2026, 5:11 PM4.3s11 files scanned
SKILL.md
- name:
- ci-fixer
- description:
- Automated CI/CD pipeline fixer - watches CI, fixes errors locally, commits, and loops until green. Use when CI is failing and you want to automatically fix and verify changes.
- argument-hint:
- [--auto] [--max-attempts=N]
- color:
- yellow
<objective>
Autonomously fix CI/CD pipeline failures by:
1. Watching CI runs in real-time
2. Analyzing failures (GitHub Actions, Vercel, Netlify, etc.)
3. Fixing errors and verifying locally (tests, linter - NO skipping, NO hacks)
4. Committing fixes and looping until pipeline is green
5. Cleaning up artifacts when done
</objective>
<quick_start> Basic usage (watch and fix CI):
/ci-fixer
Auto mode (no confirmations):
/ci-fixer --auto
With max attempts limit:
/ci-fixer --max-attempts=3
The skill will:
- Detect your current branch's CI run
- Watch for failures
- Fix errors locally (tests, lint, build)
- Commit and push
- Loop until green ✅
CRITICAL: Zero tolerance for hacks - fixes must be proper, no @ts-ignore, eslint-disable, or test skipping.
</quick_start>
<state_variables> Persist throughout all steps:
| Variable | Type | Description |
|----------|------|-------------|
| {auto_mode} | boolean | Skip confirmations |
| {max_attempts} | integer | Max fix attempts (default: 5) |
| {current_attempt} | integer | Current attempt number |
| {run_id} | string | Current GitHub Actions run ID |
| {branch} | string | Current git branch |
| {last_commit_sha} | string | SHA of commit being watched |
| {artifacts_dir} | string | Path to artifacts: .claude/data/ci-{run_id}/ |
| {error_source} | string | Source of error (github-actions, vercel, netlify) |
| {error_logs} | string | Captured error logs |
| {fixes_applied} | list | List of fixes applied this session |
| {local_verified} | boolean | Whether local tests/lint passed |
</state_variables>
<entry_point>
Load steps/step-00-init.md
</entry_point>
<step_files>
| Step | File | Description |
|------|------|-------------|
| 0 | step-00-init.md | Parse flags, detect branch, setup state |
| 1 | step-01-watch-ci.md | Find CI run, monitor status |
| 2 | step-02-analyze-errors.md | Fetch logs/artifacts, analyze errors |
| 3 | step-03-fix-locally.md | Fix errors, verify locally |
| 4 | step-04-commit-push.md | Commit and push, loop back |
| 5 | step-05-cleanup.md | Cleanup artifacts, show summary |
</step_files>
<workflow_diagram>
┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐ ┌──────────┐
│ Init │───►│ Watch │───►│ Analyze │───►│ Fix │───►│ Commit │
│ Step 0 │ │ CI │ │ Errors │ │ Locally │ │ & Push │
└──────────┘ └────┬─────┘ └──────────┘ └──────────┘ └────┬─────┘
│ │
│ ┌─────────────────────────────────────┘
│ │ (loop until green or max attempts)
│◄────────┘
│
┌────┴────┐ ┌──────────┐
│ SUCCESS │───►│ Cleanup │
│ or STOP │ │ Step 5 │
└─────────┘ └──────────┘
</workflow_diagram>
<artifacts_storage>
Artifacts are stored in .claude/data/ci-{run_id}/:
.claude/data/ci-{run_id}/
├── github/
│ ├── failed-logs.txt # Failed job logs
│ └── artifacts/ # Downloaded artifacts
├── vercel/
│ ├── deployment.json # Deployment info
│ └── logs.txt # Build/runtime logs
├── netlify/
│ └── build-logs.txt # Build logs
└── summary.md # Error analysis summary
</artifacts_storage>
<core_principles> <zero_tolerance_policy> ZERO TOLERANCE FOR HACKS
If you can't fix it properly, ASK FOR HELP. Never bypass or hack.
- NEVER skip tests - All tests must pass locally before committing
- NEVER hack around issues - Fix the root cause, don't disable checks
- NEVER use bypass flags - No
--no-verify,--skip,@ts-ignore,eslint-disable - NEVER disable rules - Don't add to
.eslintignore, don't lower strictness - Verify locally first - Always run tests/lint locally before pushing
- Minimal changes - Only fix what's broken, don't refactor unrelated code
- Clear commits - Each fix should have a descriptive commit message
- Clean up - Delete artifacts when workflow completes
Forbidden hacks include: @ts-ignore, eslint-disable, .skip, as any, --no-verify, --legacy-peer-deps, commenting out tests, changing assertions to match wrong output, adding files to ignore lists.
</zero_tolerance_policy>
</core_principles>
<success_criteria>
- CI pipeline status successfully monitored
- Errors analyzed and root cause identified
- Fixes applied locally and verified (all tests/lint passing)
- Changes committed with clear messages
- CI pipeline green ✅ after fixes
- Artifacts cleaned up
- No hacks or bypasses used
- Maximum attempts limit respected </success_criteria>
Related Skills
devattom/wiki-health-check
development
VerifiedTrustedCommunity
Use when you want to audit a project wiki for quality issues — stale version claims, contradictions between pages, orphan pages, broken wiki links, missing cross-references, or misalignment between wiki content and the actual codebase state.
SKILL.mdUpdated Apr 24, 2026
devattom/debug
development
VerifiedTrustedCommunity
Systematic error debugging with analysis, solution discovery, and verification
SKILL.mdUpdated Apr 4, 2026
devattom/workflow-debate
development
VerifiedTrustedCommunity
Structured adversarial debate between AI councillors using Agent Teams to evaluate ideas, plans, or decisions. ALWAYS use when the user says "council", "debate this", "evaluate this idea", "challenge my plan", "stress-test", "devil's advocate", "multiple perspectives", "évaluer cette idée", "débattre", "challenger mon plan", "tester cette décision", or when the user wants rigorous multi-perspective analysis of a proposal, architecture decision, or strategic choice. Each councillor (visionary, critic, pragmatist, innovator, ethicist, domain expert) represents a distinct perspective and they challenge each other through cross-examination and peer exchange, producing a nuanced verdict (PROCEED / PROCEED WITH CONDITIONS / RECONSIDER / DO NOT PROCEED). Do NOT use for divergent brainstorming or idea generation — use workflow-brainstorm instead.
SKILL.mdUpdated Apr 4, 2026
devattom/architect
development
VerifiedTrustedCommunity
Architectural refactoring - analyze codebase structure, detect architectural smells (god modules, circular deps, coupling), propose restructuring plan, safely move/rename files with git history, update imports, simplify architecture. Use when asked to "restructure", "reorganize", "simplify architecture", "move files", "split module", "fix structure", "reduce coupling", "reorganize folders", "clean up architecture", "refactor structure", "restructurer", "reorganiser", "simplifier".
SKILL.mdUpdated Apr 4, 2026