Jan 2026 – Present
Open source · Codebase intelligence for AI agents
DevLens
Analyze your repo once. Your AI agent then queries a precomputed code graph through MCP, with ranked files, call flow, impact, and security, instead of re-reading your codebase file by file.
- 50+ GitHub stars and an active community. Maintainer of devlensio/devlensOSS (AGPL-3.0), with @devlensio/cli published on npm.
- Ranked first of nine tools on every quality measure (correctness, F1, recall, precision) in a public benchmark of 2,394 agent runs against Graphify, codegraph, serena, semble, and codebase-memory.
- 67 ms per MCP call; response packets 6.7× smaller than the closest graph competitor, never exceeding a configurable token budget.
- Designed and shipped the full stack solo: multi-language parsing engine (JavaScript, TypeScript, Python, Java, Go, Rust) with 500+ tests, graph query layer, React frontend, and the @devlensio/cli npm package.
- Building DevLens Cloud on Next.js, Supabase, and Postgres: the hosted layer bringing precomputed code graphs to teams.
- LLM-powered features on the code graph: RAG-based Q&A, embeddings, and automated PR-review workflows built for both developers and agents.
- Blast-radius analysis before you change a symbol, PR review packets, and per-node security analysis.
● Actively maintained · AGPL-3.0
Stack
- TypeScript
- React
- Node.js
- Bun
- MCP
- OpenAI
- Next.js
- Supabase
- Postgres