CI Watchdog

What it does

When your agent uses it

• : - The user wants to keep an eye on CI
• Wait for checks to pass - The user mentions monitoring PR status
• GitHub checks - The user asks to watch a PR until its ready
• Fix failing checks - After pushing commits

Example prompts

• “keep an eye on CI”
• “wait for checks to pass”
• “/ci-watchdog”

Requirements

• Python 3

Workflow steps

6 steps, taken from the first numbered list in SKILL.md.

1. 1Monitor - Watch PR checks until they complete or fail
2. 2Detect - Identify which checks failed
3. 3Diagnose - Fetch logs and understand the failure
4. 4Fix - Apply appropriate fixes to the codebase
5. 5Commit - Stage, commit, and push fixes
6. 6Repeat - Go back to monitoring until all pass

What it can do on your machine

Read from SKILL.md and the folder at commit 87e8aa0. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Ships 1 file in scripts/ (Python), which the agent can run.

  Shell commands in SKILL.md call:

  • gh
  • python

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use gh, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

CI Watchdog loads about 1.6k tokens when it runs. Until then it costs about 174 tokens; SKILL.md has 455 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~174

When it runs · the whole SKILL.md, loaded when a task matches

~1.6k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); the scripts in this folder are not scanned.

SKILL.md

The full file from latitude-dev/latitude-llm at commit 87e8aa0, republished under its MIT licence (© latitude-dev). 455 words, ~1,551 tokens.

name

ci-watchdog

description

Continuously monitor GitHub PR CI checks and automatically fix failures until all checks pass. USE THIS SKILL when: - The user wants to "keep an eye on CI" or "wait for checks to pass" - The user mentions monitoring PR status, CI status, or GitHub checks - The user asks to watch a PR until it's ready or fix failing checks - After pushing commits, the user wants to ensure CI stays green - The user mentions automated fix loops for CI failures This skill handles watching GitHub Actions (or other CI) checks, detecting failures, diagnosing issues from logs, and applying fixes in a loop until all checks pass. Works with GitHub CLI (gh) and common CI systems like GitHub Actions.

# Watch current PR until all checks pass (auto-fix enabled)
gh pr checks --watch --fail-fast

# Watch with auto-fix loop
ci-watchdog --fix --loop

gh pr checks --watch

LOOP:
  1. Run: gh pr checks --watch --fail-fast
  2. IF checks pass → DONE
  3. IF checks fail:
     a. Identify failed check names
     b. Fetch logs: gh run view <run-id> --log-failed
     c. Analyze error patterns
     d. Apply fixes to codebase
     e. Commit and push
     f. GOTO LOOP

# Check failed with pyright errors in file X
python -m pyright src/path/to/file.py 2>&1 | head -30
# → Shows specific line/column errors
# → Fix the issues
# → Commit and push

# Tests failed
# Read the test file mentioned in error
# Update assertions or remove obsolete tests
# Commit and push

error: Cannot access attribute "name" for class "X"
Attribute "name" is unknown

# List current checks
gh pr checks

# Watch until complete (blocks)
gh pr checks --watch

# Watch and exit immediately on failure
gh pr checks --watch --fail-fast

# View failed check logs
gh run view <run-id> --log-failed

# View specific job
gh run view --job=<job-id>

# View current PR
gh pr view

# Push changes
gh push

# Create PR (if needed)
gh pr create --draft --title "..." --body "..."

User: "Watch this PR and fix any issues until CI passes"

→ Start: gh pr checks --watch --fail-fast
→ Wait for completion...
→ FAILED: pyright errors in latitude_span_processor.py

→ Read the failing file
→ Identify: references to removed 'name' attribute
→ Fix: Remove the name handling code
→ Commit: "fix(telemetry): remove name handling from processor"
→ Push

→ Restart: gh pr checks --watch --fail-fast
→ Wait...
→ FAILED: test failures in capture_test.py

→ Read test file
→ Identify: tests checking ctx.name which no longer exists
→ Fix: Remove ctx.name assertions
→ Commit: "test(telemetry): update tests for removed name field"
→ Push

→ Restart: gh pr checks --watch --fail-fast
→ Wait...
→ ALL PASSED ✓
→ Done!

© latitude-dev, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file (scripts) in .agents/skills/ci-watchdog of latitude-dev/latitude-llm.

• SKILL.md
• scripts/parse_failures.py

Open the folder on GitHubat commit 87e8aa0