What problem does it solve? After a feature is implemented, teams need proof that the code builds, passes mechanical checks, stays inside the locked plan's end-to-end slice, and survives an implementation-quality review before deeper behavior tests begin. This Skill runs that validation gate inside the Agentic SDLC workflow and records auditable evidence. ## Core Features & Use Cases - Mechanical Validation: Runs project-native syntax, lint, type, import, dependency, config, and build checks using the project's own configured tooling. - Locked-Slice Boundary Check: Compares changed files and entry points against the locked plan's end-to-end slice and flags out-of-scope changes. - Review-Only Code Review Gate: Invokes code-review on the current feature diff without allowing it to edit source, and records its decision in validation evidence. - Use Case: After implementing a feature in an Agentic SDLC run, the coordinator invokes this Skill to run the project's checks, verify slice boundaries, record a code-review decision in evidence/FEAT-*/validate.md, and move the feature state to validated only when everything passes. ## Quick Start Ask the Agentic SDLC coordinator to run sdlc-validate-codes for the current feature so it executes the project's build, lint, type, and config checks plus a review-only code-review before marking the feature validated.
