Architecture
The Web UI uses the same indexing pipeline as the CLI but runs entirely in WebAssembly:- Tree-sitter WASM for code parsing
- KuzuDB WASM for in-memory graph database
- transformers.js for in-browser embeddings (WebGPU/WASM)
- Sigma.js + Graphology for WebGL graph rendering
- LangChain ReAct agent for AI chat
Try it now: gitnexus.vercel.app
CLI vs Web UI Comparison
Choose the right mode for your use case:| CLI + MCP | Web UI | |
|---|---|---|
| What | Index repos locally, connect AI agents via MCP | Visual graph explorer + AI chat in browser |
| For | Daily development with Cursor, Claude Code, Windsurf, OpenCode | Quick exploration, demos, one-off analysis |
| Scale | Full repos, any size | Limited by browser memory (~5k files), or unlimited via backend mode |
| Install | npm install -g gitnexus | No install — gitnexus.vercel.app |
| Storage | KuzuDB native (fast, persistent) | KuzuDB WASM (in-memory, per session) |
| Parsing | Tree-sitter native bindings | Tree-sitter WASM |
| Privacy | Everything local, no network | Everything in-browser, no server |
Memory Limitations
The Web UI runs entirely in browser memory with the following constraints: For larger repositories:- Use the CLI + MCP for full-scale indexing
- Use local backend mode to connect the web UI to your CLI-indexed repos
Privacy & Security
100% client-side: All code processing happens in your browser. No uploads, no server-side processing.
- No network calls during indexing
- No code sent to servers — files stay in browser memory
- API keys stored in localStorage — never transmitted except to your chosen LLM provider
- Open source — audit the code yourself at github.com/abhigyanpatwari/gitnexus
- Optional LLM API requests when using AI chat (to your configured provider)
- Optional local backend connections when using bridge mode
Features
Graph Visualization
- WebGL-accelerated rendering with Sigma.js
- Interactive exploration — click nodes to see details, zoom and pan
- Community detection — functional areas highlighted with colors
- Process flows — trace execution paths through the codebase
AI Chat
- LangChain ReAct agent with graph-aware tools
- Cypher query execution — direct graph queries
- Semantic search — hybrid BM25 + vector search
- Code navigation — browse files and symbols
Search
- Hybrid search — BM25 + semantic (when embeddings are ready)
- FTS fallback — works immediately without embeddings
- Process-grouped results — findings organized by execution flows
When to Use Web UI
Best for:- Quick repository exploration
- Sharing codebase insights with non-developers
- Demos and presentations
- One-off analysis of small codebases
- When you can’t install CLI tools
- Large repositories (>5k files)
- Daily development workflow
- Production AI agent integration
- Persistent indexing across sessions