Execute

Execution means taking a scoped change all the way to a verified result. This is not "write a lot of code, then see what happens." It is a tight engineering loop: understand the slice, use the project's existing libraries and CLIs before hand-writing code, reject bandaids and code smells, prove each behaviour with TDD, keep the design healthy as the code appears, review the result, and report only evidence.

Use this skill when the user hands you a ticket, issue, plan item, acceptance criteria, or direct implementation request and expects you to build it.

The rule

One vertical behaviour at a time, verified end to end.

Do not batch the work horizontally: no "models first, then API, then UI, then tests." A completed increment should be small, coherent, and runnable: one observable behaviour, through the real public interface, with the tests and verification that prove it.

Phase 0 - Load the right companion skills

Execution is an orchestrator. Load the other skill only when its trigger applies: