What problem does it solve? After a project scan produces a value-weighted evidence baseline, teams need a deterministic way to turn that raw evidence into a queryable project cognition database without hand-writing SQL or trusting chat memory. This Skill validates scan packets, compiles them through the project-cognition binary, and publishes a schema v5 SQLite runtime with graph nodes, edges, claims, path indexes, and alias catalogs. ## Core Features & Use Cases - Deterministic graph reconstruction: Runs validate-scan, build-from-scan, and validate-build through the project-cognition CLI so the SQLite runtime is compiled from accepted evidence rather than manual SQL. - Boundary and coverage acceptance: Validates repository-universe.json, scan-targets.json, coverage ledgers, and scan packets, routing back to sp-map-scan with a gap report when evidence is incomplete. - Claim lifecycle and routing substrate: Derives revisioned graph-claim lifecycle states, alias indexes, route packs, and queryable task-oriented bundles for downstream agent workflows. - Use Case: After running sp-map-scan on a brownfield repository, invoke this Skill to publish a query-ready cognition database that later workflows can navigate with compass queries and bounded live verification. ## Quick Start Rebuild the project cognition runtime from the completed scan baseline by invoking /sp-map-build and letting it validate, compile, and publish the schema v5 database.
