Sirchmunk v0.0.2 Released — MCP Support, CLI & Knowledge Persistence

We are thrilled to announce Sirchmunk v0.0.2 — a major step forward with Model Context Protocol (MCP) integration, a comprehensive CLI, and persistent knowledge management.

Highlights

MCP Integration

Full Model Context Protocol support enables seamless integration with AI assistants like Claude Desktop and Cursor IDE. Sirchmunk’s search capabilities are exposed as MCP tools that any compliant client can discover and invoke.

{
  "mcpServers": {
    "sirchmunk": {
      "command": "sirchmunk",
      "args": ["mcp", "serve"],
      "env": {
        "SIRCHMUNK_SEARCH_PATHS": "/path/to/your_docs"
      }
    }
  }
}

The MCP server supports both stdio (for local desktop AI tools) and HTTP (for remote scenarios) transport modes.

CLI Commands

A new sirchmunk CLI provides a single entry point for all operations:

Knowledge Persistence

DuckDB-powered storage with Apache Parquet export provides efficient knowledge management:

• In-memory first: Microsecond-level read/write during runtime
• Atomic persistence: Crash-safe writes using the temp-file-then-rename pattern
• Parquet export: Columnar format for excellent compression and analytical query performance

Semantic Cluster Reuse

Knowledge clusters now store embedding vectors (384 dimensions), enabling fast cosine-similarity matching for repeated or paraphrased queries. This delivers sub-second response times for queries that match existing knowledge.

How to Upgrade

pip install --upgrade sirchmunk

Or install with all extras:

pip install "sirchmunk[all]"

What’s Next?

Our roadmap includes:

• Web search integration
• Multi-modal support (images, videos)
• Distributed search across nodes
• Knowledge visualization and deep analytics

GitHub Repository · ModelScope