Greenfield v2 application and tooling for a music / catalog knowledge graph: TypeDB + TypeQL, the gg CLI, entity-service integration, and Brave-backed search. All product and tooling development for v2 happens in this repository.
New agent? Open AGENTS.md — rules and doc index (use as Cursor / system context). Then docs/AGENT_ONBOARDING.md — project brief and read order. Handoff checklist: docs/NEXT_AGENT_TODO.md. docs/ map (active vs archive): docs/README.md. Extract stimulus (Wikipedia / MusicBrainz / Discogs + Brave): docs/WEB_ENRICHMENT.md.
docs/WORKFLOWS.md— diagrams + short prose for everyggworkflow and how TypeDB, entity-service, Brave, and canonical APIs connect.docs/GROOVEGRAPH_V2_PRODUCT_AND_BUILD_SYNTHESIS.md— canonical discovery outcomes (Q1–Q33), build defaults, and implementation-slice status (replaces the former separate Q&A log and implementer-defaults docs).
Entity-service integration (including schema pipeline: raw → validate → formatted) is described in docs/USER_AND_AGENT_GUIDE.md. Symptom matrix, pytest tags, and gg vs ES behavior: docs/AGENT_ENTITY_SERVICE_ISSUES.md.
TypeQL layout and manual apply policy: typedb/README.md. Copy .env.example to .env (gitignored) for local keys.
Python CLI and gg commands live under cli/ (uv sync, gg doctor, gg schema, gg analyze, pytest). Operator map: docs/WORKFLOWS.md.
Open groovegraph-dev.code-workspace in Cursor/VS Code if you want a named single-folder workspace for this repository.
- Prefer small, reusable modules in this repo over importing or vendoring large external trees.
- When reusing an idea or algorithm from outside, re-implement it here so ownership and licensing stay clear.