
Deeprepo
This is a local RAG engine that turns your codebase into a queryable knowledge base without external vector databases or cloud dependencies. It exposes seven MCP tools that let Claude answer questions like "how does authentication work?" or "what breaks if I change auth.py?" by pulling from a SQLite-backed code graph, semantic embeddings, and auto-generated wiki pages. The query router classifies intent and picks the cheapest retrieval strategy to cut token usage. You can run it entirely local with Ollama or mix providers like OpenAI embeddings with Anthropic's LLM. Best for teams who want codebase Q&A in their editor without shipping code to third-party services or managing infrastructure.
- tools
- 7
- last updated
- Dec 30, 2025
Tools
7 toolsPublic tool metadata for what this MCP can expose to an agent.
ingest_codebaseOne-time setup — index a repo directory
One-time setup — index a repo directory
No parameter schema in public metadata yet.
find_symbol"Where is X defined / what line is X on"
"Where is X defined / what line is X on"
No parameter schema in public metadata yet.
get_file_structure"Show me the API / functions in X"
"Show me the API / functions in X"
No parameter schema in public metadata yet.
explain_file"How does X work / explain X / what does X do"
"How does X work / explain X / what does X do"
No parameter schema in public metadata yet.
find_change_impact"What breaks if I change X"
"What breaks if I change X"
No parameter schema in public metadata yet.
ask_codebaseAny open-ended question about the code
Any open-ended question about the code
No parameter schema in public metadata yet.
get_project_overview"Give me an overview / what does this project do"
"Give me an overview / what does this project do"
No parameter schema in public metadata yet.
Sponsors
Readme
DeepRepo — Local RAG Engine for Codebases
A production-grade Python library for performing RAG (Retrieval Augmented Generation) on local codebases. No heavy frameworks, no external vector DBs, no cloud required.
What It Does
DeepRepo ingests a codebase and builds three things simultaneously:
| Layer | What it stores | Used for |
|---|---|---|
| Code Knowledge Graph | Classes, functions, imports, call edges (SQLite) | Symbol lookup, blast-radius analysis |
| Embeddings + FTS index | Semantic vectors + full-text search | Relevant code retrieval |
| Hierarchical Wiki | Plain-English .md files per module | AI explanations, chat context |
A smart query router classifies every question and picks the cheapest context strategy, reducing LLM token usage by 5–50x compared to naive RAG.
Features
Details
- Repository
- github.com/abhishek2432001/deeprepo
- Categories
- AI & LLM Tools, Search & Web Crawling
- Registry
- MCP registry, active