Project
GospeLib is a scholarly scripture study platform for Latter-day Saints. The project is in active development with 91% of milestone issues complete (332 of 365 across 31 milestones). The core web application, backend services, ingest pipeline, and plugin architecture are built. Remaining work concentrates on integration validation, desktop/mobile clients, and scholarly data acquisition.
Current Status
- 22 of 31 milestones at 100% completion
- 8 backend services operational (Go and Python)
- 14-stage ingest pipeline loading corpus data into FalkorDB
- 10 core plugins built (interlinear, witnesses, AI, commentary, journal, notes, and infrastructure plugins)
- 700+ PRs merged; 2,200+ commits on the stage branch
The primary gap is not missing code but integration and validation — many GUI features exist as isolated components that need wiring into the live application and formal acceptance testing.
Key Documents
| Document | What it covers |
|---|---|
| Development Story | Comprehensive record of what was built, how, and what was learned |
| Executive Summary | Product vision, market opportunity, and business model |
| MVP Implementation Plan | Sprint-by-sprint plan for M00-M11 milestones |
| Milestones | Detailed specifications for all 31 milestones (M00-M30) |
| UI Feature Index | Catalog of every UI feature with implementation status |
Architecture at a Glance
The system is a polyglot microservices monorepo orchestrated by Nx + pnpm workspaces:
- Frontend: Next.js 15 web app with plugin architecture, Docusaurus docs site
- Backend: Go (Gateway, Auth, Billing, Notifications) and Python (Content, AI, Ingest, Plugin Registry)
- Data: FalkorDB (graph), PostgreSQL (users/subscriptions), Redis (cache/sessions), Typesense (search)
For the full service and port listing, see the Port Map.
Development Model
The project uses a parallel batch execution model where work is decomposed into conflict-free batches across eight tracks (A through X). Each batch is one branch, one PR, one agent session. Formal phase gates validate that parallel batches compose correctly before advancing. See the Development Story for full details on this execution model.