The Agent Skills Directory

[SAFE]: The skill provides a set of procedural instructions for an AI agent to follow during feature development. It does not contain executable code, automated network requests, or unauthorized file access.
[SAFE]: The workflow incorporates multiple 'STOP' gates that require explicit user validation before proceeding to subsequent phases, ensuring human oversight of the agent's actions.
[SAFE]: Security-positive practices are emphasized, such as explicitly forbidding automatic PR merges and requiring local test suites and formatters to pass before code is committed.

ship-feature