MCP Server
MCP
Code Index
By SQLite-backed hybrid (vector + FTS) code search index for Claude Code over MCP.
Install
uvx mcp-code-indexConfiguration Example
{
"remotes": [],
"packages": [
{
"registryType": "pypi",
"identifier": "mcp-code-index",
"version": "1.0.6",
"transport": {
"type": "stdio"
},
"environmentVariables": [
{
"description": "Embedder backend. Only 'jina' (local jina-embeddings-v2-base-code via sentence-transformers) is supported; the variable exists for future expansion.",
"format": "string",
"name": "CODE_INDEX_EMBEDDER"
},
{
"description": "HuggingFace model id. Defaults to 'jinaai/jina-embeddings-v2-base-code'. Override only if you know the model is dim-compatible.",
"format": "string",
"name": "CODE_INDEX_EMBED_MODEL"
},
{
"description": "Embedding dimension. Defaults to 768 (matches jina-embeddings-v2-base-code). Must match the model's output dim.",
"format": "string",
"name": "CODE_INDEX_EMBED_DIM"
},
{
"description": "Path to the SQLite index database. Defaults to .claude/index.db relative to the working directory.",
"format": "string",
"name": "CODE_INDEX_DB"
},
{
"description": "Encode batch size (default 32). Lower (e.g. 8 or 4) to reduce peak memory if the GPU OOMs on large files.",
"format": "string",
"name": "CODE_INDEX_EMBED_BATCH"
},
{
"description": "Torch device override: 'cpu', 'mps', or 'cuda'. Unset = auto-detect. Set 'cpu' to avoid Apple-Silicon MPS out-of-memory errors.",
"format": "string",
"name": "CODE_INDEX_EMBED_DEVICE"
}
]
}
]
}
mcp
model-context-protocol
pypi
Comments
Sign in to leave a comment