A local-first code-intelligence engine for AI coding agents: hybrid search, dependency-graph impact analysis and git history over your own repo, served via MCP. We post technical deep-dives, build notes and release updates.
Contextual Laboratories is one person: Aarjun Mahule, building from India. Contextual started after most of a project's time went on re-explaining the codebase to AI tools, chat after chat. So I built one source of context for the codebase, shared by every MCP-capable agent, living on your own machine. No code ever leaves your device. It is closed source, $10 a month, with a 14-day trial and no card. On this page: technical deep-dives, build notes and release updates.
Our stack
Engine: Python, tree-sitter, local ONNX embeddings, LanceDB, served over MCP. Website: Next.js on Cloudflare. Services: Cryptolens (licensing), Clerk (sign-in), Dodo Payments (payments), Sentry (error tracking). Built mostly with Claude Code; every change passes 2,000+ tests and a 1,000+ query eval suite.
0 posts published
1 member
loading...
We're a place where coders share, stay up-to-date and grow their careers.