spec-kit-project-cognition-gate

To validate technical claims in brownfield codebases, you enforce minimal live-read constraints and prove recommendations directly from active source files rather than relying on graph JSON artifacts. This prevents speculative technical decisions during debugging and code review.

Project cognition for existing repositories is an advisory navigation mechanism that uses a cognition bundle to select minimal reads and verification routes before source grounding. It governs concept selection ranking, filtering, and carry-forward during planning and implementation workflows.

Verifying cross-repo references during codebase auditing requires checking that project cognition runtimes are fresh and ready across all referenced projects. If runtimes are stale, technical claims are blocked and you fall back to minimal live reads plus explicit map-maintenance.

You cannot use graph JSON artifacts alone to debug changes in a Spec Kit Plus repository. The skill mandates that all technical claims be proved from live files, enforcing evidence-based changes through minimal live reads instead of relying on cached artifacts.

You should not use advisory project cognition for codebase auditing when cross-project cognition runtimes are not fresh. In this state, the system blocks technical claims and requires falling back to minimal live reads plus explicit map-maintenance when requested.