Notion Prompts Server

Example Use Cases

• 1

  Fetch and list all available prompt templates stored in a Notion database to select appropriate prompts programmatically.

• 2

  Compose prompts by injecting user input and dynamic variables (current date/time, prompt name) into Notion-based templates before sending them to an LLM or external API.

• 3

  Search, filter, and retrieve prompts by category, name, or full-text search and benefit from caching to reduce Notion API calls.

Description

This server implements the Model Context Protocol (MCP) to provide access to prompt templates saved in a Notion database. It supports listing, searching, and retrieving prompts by name or ID, composing prompts with user input and time/name variables, grouping by category, and caching to reduce API calls. Configuration is via environment variables or CLI arguments and it supports multiple prompt-handling modes (return only, process locally, or call an external API). The project is open-source and distributed under the MIT license.