Enhances AI coding assistants by providing intelligent, codebase-aware context management through local hybrid RAG.
Sponsored
Ctx-Sys acts as a smart librarian for your AI coding assistant, indexing your codebase, understanding symbol relationships, and retrieving precise context. It overcomes context window limitations by combining vector search, keyword retrieval, and graph traversal with reciprocal rank fusion. Running entirely locally with Ollama, it ensures your code remains private while providing deep project understanding to any MCP-compatible AI client.
주요 기능
01Compatibility with any MCP client, exposing 12 action-based tools
02Local-first operation with Ollama for embeddings and summarization
03Semantic knowledge graph storing entities, relationships, and conversation memory
04Hybrid RAG combining vector search, keyword/FTS5, and graph traversal
05Code-aware parsing with tree-sitter AST for symbol extraction and relationships
062 GitHub stars
사용 사례
01Retrieve highly relevant code and documentation snippets for complex AI queries.
02Provide AI coding assistants with deep, codebase-wide understanding for more accurate suggestions.
03Enable privacy-focused AI code assistance with local context management and knowledge retention.
