| name | your-skill-name |
| description | [What it does — one sentence]. Use when user [trigger phrases users would actually say]. Handles [key capabilities]. Do NOT use for [negative triggers if needed to prevent over-firing]. |
| metadata | {"author":"your-name","version":"1.0.0"} |
Your Skill Name
[One paragraph: what this skill does and the outcome it produces.]
Instructions
Step 1: [First Major Step]
[Clear, specific instruction. Not "validate the data" but "Run python scripts/validate.py --input {filename} to check data format."]
- [Substep with exact command or action]
- [Substep with expected output]
Step 2: [Second Major Step]
[What to do, what tools to call, what parameters to pass.]
Expected output: [describe what success looks like]
Step 3: [Continue as needed]
[Keep steps sequential and unambiguous.]
Examples
Example 1: [Common Scenario]
User says: "[realistic trigger phrase]"
Actions:
- [What the skill does first]
- [What it does next]
Result: [What the user gets]
Example 2: [Edge Case]
User says: "[another realistic phrase]"
Actions:
- [Steps for this case]
Result: [Expected outcome]
Common Issues
[Error or problem name]
Cause: [Why it happens]
Fix: [Exact steps to resolve]
[Another common issue]
Cause: [Why]
Fix: [How]
Rules
- [Non-negotiable constraint 1]
- [Non-negotiable constraint 2]
- [Safety rails — what the skill must never do]
