Danbooru-Turso MCP Server

Simple MCP server for collecting Danbooru character data and storing it in Turso database.

🎯 Purpose

Automatically collect and save Danbooru character data to Turso cloud database with automatic upsert (updates existing posts).

✨ Features

📥 Collect character data from Danbooru API with automatic pagination
💾 Store in Turso cloud database
🔄 Auto-update existing posts (INSERT OR REPLACE)
⚡ Fast & Simple - Single tool, single purpose

🛠️ Tool

`collect_and_save`

Collect all posts for a character tag and save to Turso database.

Parameters:

tag (required): Character tag to collect (e.g., "yatogami_tenka")
max_posts (optional): Maximum number of posts to collect

Example:

{
  "tag": "yatogami_tenka",
  "max_posts": 1000
}

Response:

{
  "success": true,
  "tag": "yatogami_tenka",
  "collected": 119,

  "new_posts": 111,
  "updated_posts": 8
}

📊 Database Schema

This MCP uses the following table structure:

CREATE TABLE IF NOT EXISTS posts (
  id INTEGER PRIMARY KEY,
  tags TEXT NOT NULL,
  tag_count INTEGER NOT NULL,
  rating TEXT NOT NULL,
  score INTEGER NOT NULL,
  fav_count INTEGER NOT NULL,
  file_url TEXT NOT NULL,
  pixiv_id INTEGER
);

Note: If you need additional fields (like image_width, image_height, created_at, etc.), you can modify src/services/turso.ts to include them. The table is only created if it doesn't exist (CREATE TABLE IF NOT EXISTS), so existing data is safe.

Danbooru API returns 50+ fields per post. This MCP stores the most commonly used fields for performance and simplicity. The current schema covers the essential data for most use cases.

🚀 Deployment

Smithery (Recommended)

Push to GitHub
Deploy on Smithery with environment variables:
- TURSO_DATABASE_URL - Your Turso database URL
- TURSO_AUTH_TOKEN - Your Turso auth token

Local Development


npm install
npm run build
npm start

📦 Environment Variables

TURSO_DATABASE_URL - Turso database URL (required)
TURSO_AUTH_TOKEN - Turso auth token (required)
PORT - Server port (default: 3000)

📖 Usage Example

Collect data:

collect_and_save({ 
  tag: "yatogami_tenka",
  max_posts: 500 
})

Response:

Shows how many posts were collected
Shows how many were new vs updated
Handles duplicates automatically with UPSERT

📝 License

MIT

🛠️ Tool

collect_and_save

Collect all posts for a character tag and save to Turso database.

Parameters:

tag (required): Character tag to collect (e.g., "yatogami_tenka")

max_posts (optional): Maximum number of posts to collect

Example:

{ "tag": "yatogami_tenka", "max_posts": 1000 }

Response:

{ "success": true, "tag": "yatogami_tenka", "collected": 119, "new_posts": 111, "updated_posts": 8 }

📊 Database Schema

This MCP uses the following table structure:

CREATE TABLE IF NOT EXISTS posts ( id INTEGER PRIMARY KEY, tags TEXT NOT NULL, tag_count INTEGER NOT NULL, rating TEXT NOT NULL, score INTEGER NOT NULL, fav_count INTEGER NOT NULL, file_url TEXT NOT NULL, pixiv_id INTEGER );

Danbooru API returns 50+ fields per post. This MCP stores the most commonly used fields for performance and simplicity. The current schema covers the essential data for most use cases.