ai-driven-dev/framework/03-assert

Skill: assert

Validate that the work behaves as intended: run the project's assertions, iterating and fixing until they pass.

Actions

Run every applicable facet by default, or one when named. Coding (01) always applies; add 03 when the work has a UI and a frontend is running, the facet resolving the URL itself; run 02 only when architecture conformance is asked for. Skip a facet whose precondition is absent, with a noted reason. Ask only when the intent is genuinely ambiguous. Before running an action, read its file in actions/, not only the table or assets.

Transversal rules

• Gate: it returns a pass or fail verdict on the work.
• Fix loop: the coding and frontend facets fix and re-run until they pass. The architecture facet only reports, never fixes.
• Stop only when every selected assertion passes a final clean sweep.
• On supplied CI failures, fixing facets collect all failed checks, repair each reproducible failure locally, and rerun every applicable command in the final sweep; the caller pushes once after green if authorized, with CI-only checks unverified.

Assets

• assets/task-template.md: the tracking file the frontend facet fills across its fix loop.