d0c4b1ae28
- tests/integration-test.ts: clones p-limit repo, parses, generates diagrams (11/11 pass) - tests/pipeline-test.ts: mock LLM provider pipeline test (29/29 pass) - Fix chunkCode to handle single lines exceeding maxChars limit - Add tsx devDependency for test execution
CodeBoard
Codebase → Onboarding Docs Generator. Paste a GitHub repo URL, get interactive developer onboarding documentation in minutes.
Architecture
codeboard/
├── apps/
│ ├── web/ # Next.js 14 frontend + API routes
│ └── worker/ # BullMQ job processor
├── packages/
│ ├── shared/ # TypeScript types
│ ├── parser/ # Babel-based AST parser (JS/TS) + regex (Python)
│ ├── llm/ # OpenAI/Anthropic abstraction + prompt pipeline
│ ├── diagrams/ # Mermaid diagram generators
│ └── database/ # Prisma schema + client
Quick Start
# Install dependencies
npm install
# Build all packages
npm run build
# Start with Docker
docker compose up
Development
# Run dev server (all workspaces)
npm run dev
# Build
npm run build
Tech Stack
- Frontend: Next.js 14, React 18, Tailwind CSS 4
- Backend: BullMQ workers, Redis pub/sub for real-time progress
- Parser: @babel/parser for JS/TS, regex-based for Python
- LLM: Provider abstraction (OpenAI GPT-4o / Anthropic Claude)
- Diagrams: Mermaid.js auto-generated architecture & dependency graphs
- Database: PostgreSQL + Prisma
- Queue: Redis + BullMQ
- Deployment: Docker multi-stage builds
License
MIT
Built by Vectry — Engineering AI into your workflow.