Axon v0.1.0
π Initial Public Release
First public release of axon-rag β a local-first RAG engine for querying your documents, code, and knowledge with any LLM. Runs entirely on your hardware; no data leaves your machine.
Install
pip install axon-rag⨠Highlights
54 file formats β PDF, DOCX, PPTX, Jupyter notebooks, 24 code languages, images, URLs, and more.
Multi-LLM support β Ollama, OpenAI, Gemini, xAI Grok, GitHub Copilot API; hot-swap providers without restart.
Advanced retrieval pipeline β hybrid semantic + keyword search, HyDE, multi-query, step-back prompting, BGE reranker, RAPTOR summarisation, GraphRAG entity/relation extraction.
MCP server β 30 tools for Claude Code, Codex, Gemini CLI, Cursor, and Copilot agent mode.
VS Code extension β @axon chat participant, 3D Knowledge Graph panel, Code Graph, and Governance Console.
REST API β 54 endpoints with Swagger docs at /docs.
AxonStore β signed read-only project sharing across OS users with per-key revocation.
Entry Points
| Command | What it does |
|---|---|
axon |
Interactive REPL |
axon-api |
FastAPI REST server (port 8000) |
axon-mcp |
MCP stdio server |
axon-ui |
Streamlit UI (pip install axon-rag[ui]) |
axon-ext |
Install the bundled VS Code extension |
Documentation
Full docs at https://github.com/jyunming/Axon/tree/main/docs