MCP Apple Notes

MCP Apple Notes bridges the gap between an AI assistant and a user’s personal knowledge stored in Apple Notes. By exposing the notes database as an MCP server, Claude (or any MCP‑compliant client) can query, create, and manage notes with natural language commands. This eliminates the need to manually export or copy content into a separate knowledge base, keeping information in its native format while still enabling powerful semantic search.

The server performs a two‑stage indexing process: first it extracts the full text of each note, then it generates dense embeddings with an on‑device model. These vectors are stored in LanceDB, a fast, local vector store that scales to thousands of entries. Once indexed, the search-notes tool can return results based on meaning rather than just keyword matches, allowing an assistant to surface the most relevant notes even when the query wording differs from the original text. Full‑text search remains available for quick lookups, and list-notes provides an overview of all titles.

Key capabilities include:

• Native Apple Notes integration – no data export required; the server accesses the local Notes database directly.
• Semantic search – embeddings enable context‑aware retrieval, improving recall on ambiguous queries.
• Vector storage with LanceDB – efficient similarity search and quick updates as notes change.
• Command‑style tools – , , and let developers build workflows that automatically sync new information into the assistant’s memory.

Typical use cases are:

• A developer building a personal knowledge‑management bot that can answer questions about meeting notes, recipes, or project ideas stored in Apple Notes.
• A researcher who wants to surface related literature notes when drafting papers, without leaving the AI interface.
• An organization that uses Apple Notes as a lightweight collaboration tool and needs an assistant to surface relevant documents across teams.

Integration is straightforward: once the MCP server is running, any MCP‑enabled IDE or chat interface can invoke its tools via simple commands. The assistant can prompt the user for a query, pass it to search-notes, and then present the most relevant note excerpts, optionally creating new notes with based on follow‑up conversations. This tight coupling turns Apple Notes into an active, searchable component of the AI workflow, enhancing productivity without compromising data locality or privacy.