Documentation
get_skill
Fetch the authoritative body for a specific Milkey skill slug.
What it does
`get_skill` is the retrieval step. Call it after the agent already knows which skill it wants to use.
This tool returns the main skill body that the agent should read before it attempts the technical work the skill covers.
When to use it
- You already have a chosen skill slug from `resolve_skill` or from explicit product logic.
- The task is now specific enough that the agent should read the real guidance instead of continuing to search.
- You want a stable, authoritative instruction set before code changes or implementation planning begins.
Inputs
- slug (required): The exact skill slug returned by discovery or selected intentionally by the workflow.
- topic (optional): An optional subtopic when the skill supports narrowing to a focused subsection.
- tokens (optional): An optional token budget when the client wants to constrain how much skill content is returned.
Example request
{ "slug": "nextjs-docs-architecture", "topic": "sidebar-structure", "tokens": 5000}Response shape
- The full skill body or a focused subsection, ready for the agent to read and follow.
- Structured guidance that can then drive implementation, review, or planning steps.
In practice, the most important handling rule is:
- Use the skill body as the implementation guide, not as a suggestion to be ignored.
- If you need extra attached material after reading the main body, fetch only the specific supporting reference you actually need.
After get_skill, the next page to read is Authorization.