KVANTRA-dev/NOUZ-MCP

Overview

A Python-based MCP server that exposes local knowledge systems for querying via the Model Context Protocol. It enables AI agents to retrieve information from a locally stored knowledge base.

Best for

Best for
Developers needing a lightweight, local knowledge server for MCP-based agents.

Use cases

• Querying personal documents or notes locally
• Integrating local knowledge into MCP-compatible AI assistants
• Building a private, offline knowledge retrieval system

How to use

Install

pip install nouz-mcp

Tools exposed

• suggest_metadata
• write_file
• update_metadata
• read_file
• calibrate_cores
• recalc_signs
• recalc_core_mix
• index_all
• chunk_text
• chunk_file
• search_chunks
• list_files
• get_children
• get_parents
• suggest_parents
• add_entity
• process_orphans
• OBSIDIAN_ROOT
• NOUZ_CONFIG
• NOUZ_DATABASE_NAME

Tested with

Claude Desktop, Cursor, ChatGPT

Notes

A Python-based MCP server that exposes local knowledge systems for querying via the Model Context Protocol. It enables AI agents to retrieve information from a locally stored knowledge base.

14 stars on GitHub. Last updated 2026-05-30. Licensed MIT.

Use cases

• Querying personal documents or notes locally
• Integrating local knowledge into MCP-compatible AI assistants
• Building a private, offline knowledge retrieval system

Pros

• Simple implementation in Python
• Leverages the MCP standard for interoperability
• Keeps data local, supporting privacy

Cons

• Very limited adoption (14 stars)
• Minimal documentation and feature scope
• Only available for Python environments

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.