MCP Server: io.github.sairam0424/trelix
This MCP server provides semantic code search for AI agents using a hybrid approach that combines BM25 and vector search. It also supports call-graph–based search and covers 20+ programming languages. The intent is to retrieve relevant code artifacts beyond keyword-only matching.
🛠️ Key Features
- Semantic code search for AI agents
- Hybrid BM25 + vector retrieval
- Call-graph search
- Support for 20+ languages
🚀 Use Cases
- Finding relevant code for AI agents
- Searching across multiple programming languages
- Using relationships from call graphs to locate code paths
⚡ Developer Benefits
- More relevant results via BM25 + vector hybrid matching
- Call-graph context for locating connected code
⚠️ Limitations
- Only “semantic code search” capability is described; no other MCP tools, authentication, or indexing details are provided.