MichaelVessia/monitor-pr
modules/programs/agents/shared/skills/monitor-pr/SKILL.md
Watch an open PR's CI and bot comments; on relevant failure or new bot feedback, invoke `fix-pr`. Exit when the PR is clean or stuck. Use when user says "monitor PR", "watch CI", or wants hands-off CI shepherding without a merge step.
$ install --global
skillsauthnpx skillsauth add MichaelVessia/nixos-config monitor-prInstall 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: May 1, 2026, 10:27 AM136.6s1 file scanned
SKILL.md
- name:
- monitor-pr
- description:
- |
- allowed-tools:
- Bash(gh *), Bash(git *), Bash(sleep *), Skill
monitor-pr
Arguments: $ARGUMENTS (optional PR number or URL; else current branch PR).
Execute fully, stop only if blocked or stuck past the attempt budget.
1) Resolve target PR
- If
$ARGUMENTShas a PR URL/number, use it. Elsegh pr viewon current branch. - Capture: number, URL, head SHA, head ref.
- Abort if no PR found.
2) Monitor loop (fail-fast)
Poll checks + bot/reviewer comments together. Max 120 iterations, 60s apart (~2h). On each iteration:
- Check status rollup:
gh pr view <PR> --json statusCheckRollup --jq ' .statusCheckRollup[] | [.name, .status, .conclusion] | @tsv ' - Bot comments + reviews since last iteration:
gh pr view <PR> --json comments,reviews --jq ' [.comments[], .reviews[]] | map(select( .author.is_bot == true or (.author.login | test("bot|copilot|dependabot|renovate|coderabbit"; "i")) )) | length ' - Evaluate:
- Any check FAILED and relevant? (use
gh pr diff --name-onlyto decide relevance: touched files, lint, typecheck, build, tests tied to changed code.) Break, go to step 3. - New bot comment appeared? Break, go to step 3.
- Human review with CHANGES_REQUESTED? Break, go to step 4 (blocked).
- All checks PASSED and no new bot comments? Go to step 5 (clean).
- Checks still PENDING, nothing failed?
sleep 60, continue.
- Any check FAILED and relevant? (use
- If 120 iterations pass with checks still pending, proceed to step 4 (blocked: CI never finished).
3) Fix
Invoke fix-pr with the PR number. It handles failing checks, comments,
commits, pushes, replies.
After fix-pr returns, re-enter step 2. Max 3 total fix-pr invocations.
After the third attempt, proceed to step 4 regardless of outcome.
4) Blocked exit
Output:
## Monitor-PR: blocked
- **PR**: <URL>
- **Reason**: <changes-requested | CI-stuck | fix-pr-budget-exhausted | other>
- **Failing checks**: <list or none>
- **Open bot comments**: <count>
- **Human reviews**: <state>
- **Next step**: <what the user needs to do>
Exit non-clean. Do not merge.
5) Clean exit
Output:
## Monitor-PR: clean
- **PR**: <URL>
- **Checks**: all passed
- **Bot comments**: addressed
- **fix-pr invocations**: 0 | 1 | 2 | 3
- **Ready to merge**: yes
Caller may now invoke merge-pr. This skill does NOT merge.
Related Skills
MichaelVessia/bro
development
VerifiedTrustedCommunity
Restate the last message in plain human language, with no jargon.
2SKILL.mdUpdated Jul 23, 2026
MichaelVessia/fanout-prs
testing
VerifiedTrustedCommunity
Fan out a batch of work items into one PR each via isolated worktree agents, review every PR before it opens, then babysit all PRs through green CI and review feedback. Use when the user wants to fan out PRs, dispatch parallel PR agents over a list of items, or run a batch of independent changes as separate PRs.
2SKILL.mdUpdated Jul 10, 2026
MichaelVessia/herdr-dispatch
development
VerifiedTrustedCommunity
Dispatch user-visible coding or research agents through Herdr with Codex-Desktop-like thread ergonomics. Use when the user asks to use Herdr to spawn, dispatch, fan out, inspect, follow up with, or monitor agent workspaces/threads.
2SKILL.mdUpdated Jun 30, 2026
MichaelVessia/plannotator-visual-explainer
development
VerifiedTrustedCommunity
Generate self-contained HTML visualizations with Plannotator theming. Use for implementation plans, PR explainers, architecture diagrams, data tables, slide decks, and any visual explanation of technical concepts. Plans and PR explainers follow Plannotator's prescriptive approach; all other visual content delegates to nicobailon/visual-explainer.
2SKILL.mdUpdated Jun 12, 2026