rag-rat
Active·★ 24·MIT·Updated 2026-10-09
rag-rat is a local repository intelligence index and MCP server that provides coding agents with deep, source-anchored insights into a codebase. It builds a comprehensive code graph, integrates Git/GitHub history, and stores durable repo memories to offer provenance-backed answers about code rationale and impact.
#Repo Intelligence#Code Graph#RAG#Coding Agent Tooling#Git History Analysis#Code Duplication Detection#Semantic Search#Developer Productivity
01
Features
01Provenance-backed Answers: Every result carries confidence labels, coverage warnings, and raw evidence from the code graph, Git history, and repo memories.
02Durable Repo Memories: Source-anchored notes (Invariant, Decision, Risk) persist across refactors and sessions, surfacing automatically during queries to provide context that traditional grep cannot.
03Comprehensive Code Graph: Builds a real code graph across multiple languages (Rust, TypeScript, Python, C/C++, Swift) with optional compiler-grade resolution for enhanced accuracy.
04Git/GitHub History as Evidence: Integrates queryable Git history, lazy blame, and cached GitHub issue/PR rationale for historical context.
05Code Duplication Detection: Warns agents about exact or near-duplicate code during writing, encouraging reuse over re-implementation.
02
Compatibility
Python
Language
Verified via docs
Rust
Language/Runtime
Verified via docs
TypeScript
Language
Verified via docs
Linux
OS
Verified via docs
macOS
OS
Verified via docs
Windows
OS
Verified via docs
03
Quick start
1
$ npm install -g @rag-rat/bin
2
$ rag-rat init --yes
3
$ npx @rag-rat/skills
04
Use cases
↳Code Impact Analysis: Understand the blast radius of a code change by querying callers, callees, tests, recent commits, and associated repo memories before editing.
↳Semantic Code Search: Find specific code functionalities or concepts across source files and documentation using hybrid BM25 and vector recall.
↳Onboarding & Codebase Understanding: Identify important symbols, code clusters, and architectural decisions to quickly grasp a new repository.
↳Preventing Technical Debt: Detect and prevent code duplication during development and record durable invariants or decisions for future agents.
↳Maintaining Codebase Quality: Use `dream` to find coverage gaps in repo memories, stale references, and verify memory verdicts against current code reality.
05
Alternatives
mcp-local-rag★ 416
Local-first RAG server for developers. Semantic + keyword search for code and technical docs. Works with MCP or CLI. Fully private, zero setup.
docs-mcp-server★ 1.8k
Grounded Docs MCP Server: Open-Source Alternative to Context7, Nia, and Ref.Tools
ai-workshop★ 60
Building GenAI Apps in C#: AI Templates, GitHub Models, Azure OpenAI & More
solon-ai★ 466
Java AI application development framework (supports LLM-tool,skill; RAG; MCP; Agent-ReAct,Team-Agent). Compatible with java8 ~ java25. It can also be embedded in SpringBoot, jFinal, Vert.x, Quarkus, and other frameworks.
root-kg★ 16
Your knowledge graph. Ask questions across all your notes, meetings, and emails — cited answers.
Comments
Log in to leave a comment
No comments yet. Be the first!