skilldoc — Skill Doctor for OpenClaw
Scans all installed OpenClaw skills, validates their dependencies (binaries, OS requirements), and reports which skills are healthy, broken, or missing dependencies.
Usage
# Full health check
python3 skilldoc.py check
# Verbose (show healthy skills too)
python3 skilldoc.py check -v
# JSON output (for cron/heartbeat integration)
python3 skilldoc.py check --json
# List all skills with status
python3 skilldoc.py list
# Get fix commands for broken skills
python3 skilldoc.py fix-hints
# Check a specific skill
python3 skilldoc.py check --skill discord-voice
# Add extra skill directories
python3 skilldoc.py check --dirs ~/clawd/skills/ /other/path/
Features
- Scans builtin + custom skill directories automatically
- Auto-detects skill paths from
~/.openclaw/openclaw.jsonconfig - Validates required binaries via
whichchecks - OS compatibility checking
- Install hints from skill frontmatter metadata
- JSON output for automation (cron jobs, heartbeat checks)
- Works with or without PyYAML (graceful fallback parser)
Integration
Heartbeat check
python3 skilldoc.py check --json > /tmp/skilldoc-report.json
Parse JSON and alert on newly broken skills.
Cron job
Schedule periodic scans to catch breakage after updates.
Skill Dependency Declaration
For skilldoc to validate your skill's dependencies, add this to your SKILL.md frontmatter:
---
metadata:
openclaw:
os: [darwin, linux]
requires:
bins: [ffmpeg, yt-dlp]
install:
- kind: brew
formula: ffmpeg
- kind: pip
package: yt-dlp
---