implement

This is a very thin, manually-invoked skill (a Chinese-language translation of Matt Pocock's skills, with model auto-invocation disabled) that describes a disciplined implementation workflow driven by a spec or a set of tickets. It solves the problem of ad-hoc implementation by prescribing a short, opinionated routine rather than adding new tooling of its own.

The workflow it lays out is: implement the work described in the spec or tickets; use `/tdd` at pre-agreed seams wherever possible; run type-checking regularly and run individual test files regularly, then run the full test suite at the end; review the completed work with `/code-review`; and commit the work to the current branch. It is essentially an orchestration checklist that leans on sibling slash-command skills (`/tdd`, `/code-review`) for the actual heavy lifting.

Target users are developers and AI agents who have a spec or tickets ready and want a consistent, test-oriented execution loop that ends in a reviewed commit. Because the SKILL.md is only a handful of lines and delegates to other commands, its value depends on those companion skills and on the user having a spec or tickets to implement against.