name: [skill-name] # Max 64 chars, lowercase description: [1-2 sentences explaining what capability this adds. This is what the agent reads to decide if it should trigger the skill.] # Max 1024 chars
[Skill Title]
When invoked or deemed necessary, follow these exact instructions to [do the thing].
1. Analysis Phase
- Review the current state of [target area].
- Identify [key components].
2. Execution Phase
- Steps to perform.
- If a script is bundled, direct the agent:
Execute ./scripts/[script-name].sh
3. Verification Phase
- Confirm [outcome].
Note: For deep technical context, instruct the agent to read
./references/DETAILS.mdrather than polluting this root file.