OPEN LINK ↗
// 79d agoPRODUCT LAUNCH
codebase-context indexes team patterns for AI agents
A local-first Model Context Protocol (MCP) server that indexes your codebase to teach AI agents your team's specific coding conventions. It uses hybrid RRF search and tree-sitter AST extraction to provide agents with "preflight checks," identifying golden examples and declining patterns before they write a single line of code.
// ANALYSIS
While most RAG tools focus on what code exists, codebase-context focuses on how your team actually writes it, preventing AI from hallucinating "technically correct" but non-idiomatic solutions.
- –Hybrid RRF search combines semantic embeddings with keyword matching to catch specific symbols that vector-only search often misses
- –Automatic pattern detection tracks "Rising" vs. "Declining" conventions based on git recency, steering agents away from legacy debt
- –Local-first architecture using LanceDB and Xenova embeddings ensures zero-latency, private indexing without external API costs
- –The "edit-readiness gate" explicitly tells agents when their context is too thin to trust, reducing the risk of confident but wrong refactors
// TAGS
codebase-contextdevtoolmcpai-codingideopen-sourcesearch
DISCOVERED
79d ago
2026-03-08
PUBLISHED
82d ago
2026-03-05
RELEVANCE
9/ 10
AUTHOR
SensioSolar