REPLACE: 150-250 char description. Lead with 'Use when ...', include 3-5 trigger phrases the user might say, and end with what this skill produces. This text is ALWAYS in context once the plugin is enabled — keep it tight.
How this skill is triggered — by the user, by Claude, or both
Slash command
/REPLACE-WITH-PLUGIN-NAME:my-workflowThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Short orchestrator. Routes to the right stage file based on the `args` value. Bulky content lives in `stages/` and is loaded only when needed.
Short orchestrator. Routes to the right stage file based on the args value. Bulky content lives in stages/ and is loaded only when needed.
args value passed to this skill invocation. Extract the stage name.args is empty or ambiguous, ask the user which stage they need. Match these triggers:
stages/<stage>.md (relative to this SKILL.md location):
example-stage → stages/example-stage.md| Stage | When to Use | Output |
|---|---|---|
example-stage | REPLACE: one-line trigger description | REPLACE: artifact path |
REPLACE with any cross-cutting principles that apply to all stages. Stage-specific guidance belongs inside individual stage files.
Creates, edits, and optimizes skills for Claude Code, including drafting, evaluating with test prompts, iterating on performance, and improving skill descriptions for better triggering accuracy.
npx claudepluginhub nazarkalytiuk/claude-plugin-template --plugin REPLACE-WITH-PLUGIN-NAME