| name | sync-agent-configs |
| description | Sync skill definitions to agent proxy files (.agent/workflows/skills.md and .github/copilot-instructions.md). Use after adding, removing, or modifying skills. |
Sync Agent Configurations
Overview
Synchronize skill definitions from .claude/skills/ to the agent proxy configuration files. Run this after any changes to skills to keep all agent configurations in sync.
Target Files
| File | Purpose |
|---|---|
.agent/workflows/skills.md |
Generic agents, Aider, custom tools |
.github/copilot-instructions.md |
GitHub Copilot |
Instructions
Step 1: Scan All Skills
Read all SKILL.md files from .claude/skills/*/SKILL.md and extract:
namefrom frontmatterdescriptionfrom frontmatter- Key command from the
## Commandsection (if present)
Step 2: Update Available Skills Table
In both proxy files, update the "Available Skills" table with format:
| Command | Skill Location | Purpose |
|---------|----------------|---------|
| `skill: {name}` | `.claude/skills/{name}/SKILL.md` | {description (truncated to ~60 chars)} |
Step 3: Update Quick Reference Section
In .agent/workflows/skills.md, update the "Quick Reference (Inline Commands)" table:
| Skill | Command |
|-------|---------|
| {name} | `{primary command from skill}` |
In .github/copilot-instructions.md, update the "Quick Reference (If Unable to Read Skill Files)" section with individual headings and bash blocks for each skill.
Step 4: Verify Collections
Check data/ for folders with videos.csv and update the Collections table in .github/copilot-instructions.md if new collections exist.
Step 5: Report Changes
Summarize what was updated:
- Skills added
- Skills removed
- Skills modified
- Collections updated
Example Output
Synced agent configurations:
Skills (5):
✓ extract-videos
✓ download-transcripts
✓ add-video-to-collection
✓ sync-all-channels
✓ download-all-transcripts
Updated files:
✓ .agent/workflows/skills.md
✓ .github/copilot-instructions.md
Collections detected: library-of-minds, random, aws-reinvent-2025
Notes
- This skill is self-referential — it maintains the system that enables skills
- Run after any skill changes to keep proxy files accurate
- The proxy files contain inline commands as fallback for agents that can't read files dynamically