TiddlyWiki

You provision an MCP server that lets AI assistants access and manage your TiddlyWiki wikis via HTTP, with optional semantic search to find relevant tiddlers across content. This enables natural-language queries, precise filtering, and safe content creation or modification through stable, stateless interactions.

How to use

You connect an MCP client to the server and send tool requests to search, create, update, or delete tiddlers. You can perform traditional filter-based searches using TiddlyWiki syntax, request semantic search when Ollama is available, or combine both with a hybrid approach that first filters results and then applies semantic ranking.

How to install

Prerequisites you need before starting:

• Node.js 22+ installed on your system.

• TiddlyWiki with HTTP API enabled so the MCP server can talk to it.

Optional but recommended: Ollama for semantic search embeddings.

Step 1. Install and run TiddlyWiki with HTTP API enabled.

# Install TiddlyWiki globally if you haven't already
npm install -g tiddlywiki

# Create a new wiki and start it with HTTP API enabled
# Initialize a wiki named mywiki and launch its HTTP API
tiddlywiki mywiki --init server

tiddlywiki mywiki --listen port=8080

Step 2. Start the MCP server (choose one method):

# Method A: run via npx (recommended for quick start)
TIDDLYWIKI_URL=http://localhost:8080 npx tiddlywiki-mcp-server

# Method B: install globally and run the server
npm install -g tiddlywiki-mcp-server
TIDDLYWIKI_URL=http://localhost:8080 tiddlywiki-mcp-server

Step 3. (Optional) Set up Ollama for semantic search and pull the embedding model.

# Install Ollama from the official site and then pull the embeddings model
ollama pull nomic-embed-text

Additional configuration and usage notes

All configuration is done via environment variables. The most important are the TiddlyWiki URL and, if you enable semantic search, Ollama-related settings. You can run in stdio mode for desktop integrations or in HTTP mode to expose a conventional HTTP API.

Key environment variables you will encounter include the following. Values shown are examples or placeholders.

• TIDDLYWIKI_URL: URL of your TiddlyWiki HTTP API (for example, http://localhost:8080)

• MCP_TRANSPORT: Transport mode, either stdio or http

• MCP_PORT: Port for the HTTP server when using HTTP transport

• OLLAMA_URL: Ollama API URL

• OLLAMA_MODEL: Embedding model name, e.g., nomic-embed-text

• EMBEDDINGS_ENABLED: Enable or disable semantic search

• EMBEDDINGS_DB_PATH: Path to the embeddings SQLite database

• AUTH_HEADER: HTTP header used for authentication by the TiddlyWiki API

• AUTH_USER: Username for TiddlyWiki API requests

Configuration overview

All configuration is driven by environment variables. You can tailor which features are active, such as enabling semantic search and selecting the embeddings model. The server will gracefully degrade if semantic search is unavailable.

Usage examples for common tasks

Search tiddlers using a filter, semantic similarity, or a hybrid approach. You can combine a TiddlyWiki filter with a semantic query for more relevant results.

Available tools

search_tiddlers

Search tiddlers using TiddlyWiki filter syntax, semantic similarity, or a hybrid approach that combines both methods.

create_tiddler

Create new tiddlers with custom fields and content.

update_tiddler

Update existing tiddlers with an optional diff preview to show changes.

delete_tiddler

Delete tiddlers with a content preview to confirm before removal.