Quick Start

A working setup needs the package, a folder, an MCP client, and an embedding endpoint. The MCP process can start before model availability is checked, but index_all, semantic search, and bridges call the embedding API.

Install

python -m pip install linza-mcp

For direct PDF reading:

python -m pip install "linza-mcp[pdf]"

Without the optional PDF dependency, LINZA still reads Markdown, TXT, JSON, DOCX, and XLSX.

Prepare Embeddings

Local path through LM Studio:

1. Open LM Studio.
2. Download an embedding model.
3. Start Local Server.
4. Check the endpoint: http://127.0.0.1:1234/v1.

If you change provider, model, or dimension, run a full reindex.

Connect MCP

Replace /absolute/path/to/your-folder with the working folder. Without LINZA_VAULT, the server uses ./vault; for real work, set the path explicitly.

{
  "mcpServers": {
    "linza": {
      "command": "linza-mcp",
      "env": {
        "LINZA_VAULT": "/absolute/path/to/your-folder",
        "LINZA_EMBED_PROVIDER": "lmstudio",
        "LINZA_EMBED_URL": "http://127.0.0.1:1234/v1",
        "LINZA_EMBED_MODEL": "your-embedding-model-name"
      }
    }
  }
}

VS Code / Copilot MCP usually uses the top-level key servers and the field "type": "stdio".

Variables

Variable	Value
LINZA_VAULT	Working folder; defaults to ./vault
LINZA_EMBED_PROVIDER	lmstudio, ollama, or openai
LINZA_EMBED_URL	Embedding API URL
LINZA_EMBED_MODEL	Embedding model name
LINZA_EMBED_KEY	Endpoint key, if needed
LINZA_TOOL_SURFACE	default or advanced

Check

linza-mcp --version

After connecting, ask the agent:

Check LINZA through agent_workspace(action="doctor").
Run index_all for the working folder.
Show agent_workspace(action="map") and the first review proposals.
Do not apply anything without a dry-run and exact IDs.