Nexus AI — Graph-Grounded Code Intelligence
Ask questions about your codebase in plain English. Get grounded, citation-backed answers — streamed live, right inside VS Code.
Nexus AI builds a call graph + vector index of your codebase and uses it to answer questions with full structural awareness. It doesn't just search for keywords — it understands how your code is connected.
100% local & private. Your code never leaves your machine. The index lives in .nexus/graph.db inside your workspace. No cloud database, no telemetry, no server to manage.
What You Can Do
| Mode |
Description |
| Explain |
Ask anything about your code — get streamed answers with clickable file and line citations |
| Debug |
Point at a function — get a ranked list of suspects with anomaly scores and a root-cause diagnosis |
| Review |
Get structured findings (severity · category · suggestion) ready to post directly to a GitHub PR |
| Test |
Generate framework-aware tests written directly into your repo |
| Auto |
Let Nexus classify your intent and route to the right mode automatically |
Features
- Live streaming answers — tokens stream into the chat panel as the LLM generates them
- Clickable citations — every answer links to the exact file and line number in your editor
- Graph-aware retrieval — BFS call-graph traversal surfaces structurally connected code that pure vector search misses (+13% RAGAS score vs vector-only)
- Incremental re-index — file saves trigger automatic background re-indexing with a 2s debounce
- Secure API key storage — keys stored in VS Code SecretStorage (OS keychain), never written to disk or
settings.json
- GitHub PR integration — post Review findings as inline comments directly to a pull request
- Embedding mismatch detection — warns and blocks chat if you switch embedding models without re-indexing
- Multi-workspace support — each workspace gets its own isolated
.nexus/graph.db index
- Dev-mode passthrough — if port 8000 is already occupied, the extension skips spawning its own backend
Zero Setup
Install the extension — that's it. The backend binary is bundled inside the extension and starts automatically when VS Code opens your workspace. No Python installation, no terminal commands, no Docker, no configuration files required.
Getting Started
1. Install
Search Nexus AI in the VS Code Extensions panel and click Install.
2. Set your API key
Cmd+Shift+P → Nexus: Set API Key → pick your provider → paste your key.
3. Choose your provider
Code → Settings → Extensions → Nexus AI — pick chat and embedding provider.
4. Index your workspace
Cmd+Shift+P → Nexus: Index Workspace
Chat unlocks once indexing completes. Your index is saved in .nexus/graph.db inside the workspace.
5. Ask a question
Click the Nexus AI icon in the Activity Bar and start chatting.
Supported Providers
| Provider |
Chat |
Embeddings |
| OpenAI |
✓ |
✓ |
| Mistral |
✓ |
✓ |
| Anthropic |
✓ |
— |
| Google Gemini |
✓ |
✓ |
| Ollama (local) |
✓ |
✓ |
Mix and match — use Anthropic for chat and Mistral for embeddings, for example.
Changing embedding provider or model requires a re-index. Nexus will warn you and disable chat until the new index is ready.
Settings
| Setting |
Default |
Description |
nexus.chatProvider |
mistral |
LLM provider for chat |
nexus.chatModel |
mistral-small-latest |
Chat model name |
nexus.embeddingProvider |
mistral |
Embedding provider |
nexus.embeddingModel |
mistral-embed |
Embedding model name |
nexus.hopDepth |
1 |
Graph traversal depth |
nexus.maxNodes |
10 |
Max context nodes per query |
nexus.ollamaBaseUrl |
http://localhost:11434 |
Ollama base URL |
Commands
| Command |
Description |
Nexus: Index Workspace |
Build or refresh the code index |
Nexus: Clear Index |
Remove the index for the current workspace |
Nexus: Set API Key |
Store an API key securely |
Nexus: Clear API Key |
Remove a stored API key |
Source & License
