nextcloud-mcp-server

What it is and what it does

This package is a standalone MCP server that runs independently of Nextcloud (via Docker, VM, Kubernetes, or locally) and bridges AI assistants to Nextcloud data. It exposes MCP tools covering Notes, Calendar, Contacts, Files, Deck, Cookbook, Tables, Sharing, News, Mail, Collectives, and Talk apps, allowing large language models to create, read, update, and delete content through natural language conversations.

The server handles authentication via app-specific passwords (Basic Auth in single-user mode, OAuth-based Login Flow v2 for multi-user deployments) and includes optional experimental semantic search powered by vector embeddings. It supports multiple transports (streamable-http and stdio), document processing with OCR, and tag-based file exclusion for sensitive content. Deployment is flexible: run it locally with uvx, in Docker, or on Kubernetes.

Use it for:

• Enable Claude or other LLMs to create and manage Nextcloud notes and documents through conversational prompts
• Let AI assistants read and write calendar events, todos, and manage attendees in Nextcloud Calendar
• Use semantic search to find Nextcloud notes, files, or emails by meaning rather than keywords
• Integrate Nextcloud file management into IDE-based AI coding assistants for document access and organization
• Build multi-user hosted services where clients authenticate via OAuth and the server manages per-user Nextcloud access

Worth the install?

AI-flagged interpretation of the facts on this page — verify before relying

Provides a Model Context Protocol (MCP) server that connects AI assistants like Claude to a Nextcloud instance, enabling them to read and write notes, calendar events, contacts, files, and other Nextcloud app data through a unified API.

Yes, if you run Nextcloud and want to give AI assistants (Claude, GPT, Gemini) structured access to your data. The package is actively maintained, has low install friction, and covers a broad set of Nextcloud apps. AGPL-3.0-only licensing requires that any modifications be shared under the same license. No known vulnerabilities. Start with the uvx quick-start or Docker deployment; semantic search is optional and requires additional infrastructure.

Install

nextcloud-mcp-server on PyPI

pip install nextcloud-mcp-server

uv add nextcloud-mcp-server

poetry add nextcloud-mcp-server

Installing nextcloud-mcp-server

Before you install

Low install friction; pure Python wheel. Active maintenance with recent release. Requires 42 runtime dependencies including httpx, pydantic, uvicorn, and optional semantic search infrastructure (qdrant-client, fastembed). Supports Python 3.11+.

License in practice

AGPL-3.0-only: any modifications or derivative works must be released under the same license and made available to users. Suitable for open-source deployments; review required if you plan to modify and distribute privately.

Quickstart

# Quick start with uvx
NEXTCLOUD_HOST=https://your.nextcloud.instance.com \
NEXTCLOUD_USERNAME=your_username \
NEXTCLOUD_PASSWORD=your_app_password \
  uvx nextcloud-mcp-server run --transport stdio

# Or via Docker
docker run -p 127.0.0.1:8000:8000 --rm \
  -e NEXTCLOUD_HOST=https://your.nextcloud.instance.com \
  -e NEXTCLOUD_USERNAME=your_username \
  -e NEXTCLOUD_PASSWORD=your_app_password \
  ghcr.io/cbcoutinho/nextcloud-mcp-server:latest

Requires Python 3.11 or later. Nextcloud instance URL, username, and app-specific password (not login password) must be provided via environment variables.

Verify before relying

• Whether all 110+ MCP tools mentioned in the description are fully functional in this release
• Performance characteristics and latency for semantic search queries at scale
• Compatibility with specific Nextcloud server versions beyond the general 'stock Nextcloud' claim

Package facts

Evidence: nextcloud_mcp_server-0.173.1-py3-none-any.whl

Keywords: nextcloud, mcp, model-context-protocol, llm, ai, claude, webdav, caldav, carddav

Tags

More WWW/HTTP packages

Further reading