maximebaudette/agent-review
archive/skills/agent-review/SKILL.md
Weekly or on-demand audit of all agent workspaces (Cooper, Andy, MARS). Checks for stale script references, wrong skill paths, snapshot count drift, config vs file drift, and contradictory instructions. Auto-fixes simple unambiguous issues. Emails Maxime only for items requiring judgment. Silent on clean runs. MARS-only — do not invoke as Cooper or Andy.
testing
Updated May 31, 2026
$ install --global
skillsauthnpx skillsauth add maximebaudette/agent-skills agent-reviewInstall 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 26, 2026, 7:10 AM29.6s1 file scanned
SKILL.md
- name:
- agent-review
- description:
- Weekly or on-demand audit of all agent workspaces (Cooper, Andy, MARS). Checks for stale script references, wrong skill paths, snapshot count drift, config vs file drift, and contradictory instructions. Auto-fixes simple unambiguous issues. Emails Maxime only for items requiring judgment. Silent on clean runs. MARS-only — do not invoke as Cooper or Andy.
- homepage:
- https://github.com/MaximeBaudette/agent-skills
Skill: agent-review
Invoked on-demand or weekly (Mondays). MARS only — do not run this as Cooper or Andy.
Reviews all three agent workspaces (Cooper, Andy, MARS) for internal discrepancies, stale references, and incoherence. Auto-fixes simple issues. Emails Maxime only if there are items that require his judgment.
Step 1 — Read all key files for each agent
For each workspace, read the files listed below. Read in parallel where possible.
Cooper (~/.openclaw/agent_health/)
AGENTS.md,SOUL.md,TOOLS.md,HEARTBEAT.md,MEMORY.mdscripts/submit_task.py,scripts/batch_poll.py,scripts/scan_memory.py,scripts/update_watermark.py,scripts/check_symptom_followup.pyskills/lab_results/SKILL.md,skills/symptom-tracker/SKILL.md
Andy (~/.openclaw/career-manager/)
AGENTS.md,SOUL.md,TOOLS.md,MEMORY.md,WEEKLY_HUNT.md,SKILLS.mdcareer/profile/PROFILE.md
MARS (~/.openclaw/workspace/)
AGENTS.md,SOUL.md,IDENTITY.md,USER.md,TOOLS.md,MEMORY.md,HEARTBEAT.md~/.openclaw/openclaw.json(agents section only — for config vs. file drift)
Step 2 — Check for discrepancies
For each workspace, look for:
- Stale script names — does any doc reference a script that no longer exists in
scripts/? - Wrong skill paths — do AGENTS.md / TOOLS.md reference skill files that don't exist?
- Retired skills still referenced as active — e.g. deprecated skills still in the main skill list
- Snapshot count drift — does AGENTS.md / MEMORY.md say the right number of snapshot files?
- Config vs. file drift — does
openclaw.jsonreference files that are missing, or files exist that aren't registered? - Contradictory instructions — same instruction in two files says opposite things
- Stale date references — files referencing years or dates that are clearly wrong for the context
- Missing cross-references — HEARTBEAT.md references a skill that has no SKILL.md
- Broken cron/heartbeat registration — a cron job references a script that doesn't exist
Step 3 — Auto-fix what is simple and unambiguous
Fix directly (no confirmation needed) if ALL of these are true:
- The correct value is obvious (e.g., wrong filename in a path that clearly maps to one existing file)
- The fix is a string replacement or small addition — not a logic or policy change
- The fix affects only one workspace's documentation (not agent behavior)
Examples of safe auto-fixes:
batch_submit.py→submit_task.pyin a doc (script was renamed)health-snapshot/SKILL.md→lab_results/SKILL.mdin a skill list3 snapshot files→5 snapshot fileswhere the actual count is clearly 5- A typo in a file path that clearly maps to one real path
Never auto-fix:
- Agent behavior rules (SOUL.md, core AGENTS.md rules)
- Anything that changes what an agent will DO, not just how a path is named
- Items where two interpretations are plausible
For each auto-fix applied, record: [file] line X: changed "old" → "new"
Step 4 — Compile items requiring judgment
Collect all discrepancies that were NOT auto-fixed. For each:
- File path and line (or section)
- What the problem is
- What the options are (if not obvious)
Step 5 — Output
If there are judgment-required items:
Send an email to [email protected]:
Subject: [Agent Review] X items need your attention — YYYY-MM-DD
Body:
Hi Maxime,
Weekly agent review complete.
Auto-fixed:
- [list of auto-fixes, or "none"]
Items needing your decision:
1. [Cooper / Andy / MARS] — [file] — [description]
Options: [a] ... [b] ...
2. ...
Let me know how you'd like to handle each.
— MARS
If everything is clean (no judgment items):
Do nothing. No email, no Telegram message. Silent success.
If auto-fixes were made but no judgment items:
No email needed. Optionally log to ~/.openclaw/workspace/memory/YYYY-MM-DD.md:
## [HH:MM PT] Agent review — auto-fixed X items, all clean
[list of fixes]
What this skill does NOT do
- Does not modify agent behavior (SOUL, core identity, behavior rules)
- Does not touch session files, credentials, or gateway config
- Does not check Andy's application pipeline or career data (out of scope)
- Does not run Cooper's heartbeat or submit any health batches
Related Skills
maximebaudette/latex
development
VerifiedTrustedCommunity
Compile LaTeX documents using the Tectonic engine. Use this skill whenever the user wants to create, edit, or compile a LaTeX document — letters, articles, CVs, reports. Tectonic is a self-contained compiler that auto-downloads only the packages it needs.
SKILL.mdUpdated May 31, 2026
maximebaudette/stack-summary
tools
VerifiedTrustedCommunity
Maintain living stack documentation for the MARS host in the Prime Radiant KB. Use this skill whenever: updating the mars host page after infrastructure changes, appending archive entries for deprecated components, documenting a new service/skill/plugin being added or removed, or when the user asks to 'update stack docs', 'document this change', 'archive this setup', 'what's in the stack', or 'generate stack summary'. Also use when you've just finished installing or removing any service, plugin, skill, tool, or cron — proactively suggest documenting it.
SKILL.mdUpdated May 31, 2026
maximebaudette/gemini-cli
tools
VerifiedTrustedCommunity
Run autonomous dev tasks using Gemini CLI (gemini-cli) with Maxime's free Google quota. Use for programming projects where MARS needs to delegate agentic coding work — file edits, code gen, refactoring, analysis — to Gemini in a project directory.
SKILL.mdUpdated May 31, 2026
maximebaudette/agent-dispatch
documentation
VerifiedTrustedCommunity
Reusable inter-agent dispatch contract; responsible for delivering structured handoff requests from MARS to other sub-agents (Andy, Cooper) via the documented transport path.
SKILL.mdUpdated May 31, 2026