my-skill-name

name	my-skill-name
description	TODO — What this skill does + when to use it, with the trigger keywords a user would actually type. Required: max 1024 characters, third person. Example: "Generate API client stubs from an OpenAPI spec. Use when the user mentions OpenAPI, swagger, API clients, or asks to generate SDK code."

My Skill Name

TODO — one paragraph: what this skill does and what "done" looks like.

Remove this note and every TODO before shipping. Delete any section you don't need — an unused section is worse than a missing one, because the agent will follow it.

Workflow

TODO — numbered steps with the expected outcome of each:

1. TODO — first step (e.g. run a command, read a file).
2. TODO — the decision or generation step.
3. TODO — validation step. Prefer a script: python scripts/example.py <artifact> Fix reported violations and re-run until exit 0.

Rules

• TODO — boundaries: what this skill must never do.

Examples

TODO — one realistic input → expected output pair. One example beats a paragraph of description.

Reference files

• references/REFERENCE.md — read when TODO (state the condition, e.g. "the spec has non-default auth").
• assets/example_template.md — render final output into this structure; keep section order exactly.

Scripts

• scripts/example.py — TODO: what it checks, its exit codes, and what to do when it fails.