pin-it-mcp
Read-only stdio MCP server that allows AI clients to list, search, and retrieve pins and images from a local desktop pinboard.
README
<p align="center"> <img src="build/icon.svg" alt="Pin It" width="88" height="88"> </p>
Pin It
Pin it, so your AI can use it later.
Pin It is a local desktop pinboard for saving text, tables, code, links, and screenshots while you work. When something becomes useful again, you can copy it directly or bring it into an MCP-capable AI client.
[!IMPORTANT] Pin It is an early macOS-first project. There is not yet a signed, notarized public installer. Running from source is currently the supported installation path. Clearly labeled unsigned community builds may be added later, but macOS can warn before opening them and may require permissions to be granted again after an update.
What it does
- Pin the clipboard with
Command+Shift+I. - Pin a screen region with
Command+Shift+O. - Organize pins into multiple local boards.
- Save text, Markdown, code, HTML tables, links, and images.
- Search, reorder, edit, archive, and delete pins.
- Copy selected material as a deterministic Markdown packet.
- Read pins from MCP-capable clients through the bundled read-only server.
Privacy
Pins and images stay on your computer. Pin It has no account system, cloud sync, telemetry, or remote pin storage.
Smart Titles are optional. If you enable them and configure an OpenAI API key, Pin It sends a capped excerpt of newly captured text to OpenAI after the pin is saved locally. Without a key, Pin It uses local fallback titles. Image pins are not sent for Smart Titles.
The API key is stored in macOS Keychain. On other supported Electron platforms, Pin It uses Electron's encrypted storage when it is available.
Requirements
- macOS for the complete capture experience
- Node.js 22 or newer
- npm
Run from source
git clone https://github.com/johnnyzhoujz/pin-it-public.git
cd pin-it-public
npm ci
npm run desktop
The renderer is also available on http://127.0.0.1:5173/ while the desktop
development process is running.
To run the Electron shell against production-built renderer assets:
npm run desktop:prod
Local data
Pin data is stored under Electron's app-data directory. The directory still uses the project's former name so existing local data survives the rename:
~/Library/Application Support/Keep That/keeps.json
~/Library/Application Support/Keep That/images/
keeps.json stores relative image paths and pinit://pin/<id>/image
placeholders rather than inline base64 payloads.
MCP
Pin It includes a read-only stdio MCP server. It can list and search pins, retrieve image pins, and generate deterministic packets.
Build and test the server:
npm run build:mcp
node dist-mcp/pinit-mcp.mjs --selftest
Example configuration for a repository checkout:
{
"mcpServers": {
"pin-it": {
"command": "<absolute-path-to-node>",
"args": ["<absolute-path-to-repo>/dist-mcp/pinit-mcp.mjs"],
"env": {
"PINIT_STORE_PATH": "~/Library/Application Support/Keep That/keeps.json"
}
}
}
}
Example configuration for a locally packaged app:
{
"mcpServers": {
"pin-it": {
"command": "/Applications/Pin It.app/Contents/MacOS/Pin It",
"args": [
"/Applications/Pin It.app/Contents/Resources/app.asar.unpacked/dist-mcp/pinit-mcp.mjs"
],
"env": {
"ELECTRON_RUN_AS_NODE": "1",
"PINIT_STORE_PATH": "~/Library/Application Support/Keep That/keeps.json"
}
}
}
}
The Settings panel shows the resolved configuration for the current app/session.
Smart Titles
Smart Titles are opt-in. Add an OpenAI API key in Settings, or provide it only to the development process:
OPENAI_API_KEY=... npm run desktop
Captured content is saved immediately with a local title before any provider request runs. Provider failures leave the local title unchanged, and a generated title never overwrites a title you edited manually.
Validate changes
Run the complete local check:
npm run check
Or run the steps separately:
npm run audit:prod
npm test
npm run build
Create a local macOS build
Create an unpacked development app:
npm run pack:mac
Create local .dmg and .zip artifacts:
npm run dist:mac
Generated artifacts are written to release/ and are excluded from Git.
Unless you configure a valid Apple Developer ID identity and notarization,
these builds are for local testing and are not equivalent to an official
signed release.
Project layout
electron/ Electron main process, preload bridge, updater, and MCP entrypoint
src/ React renderer and local pinboard behavior
tests/ Node and Vitest regression coverage
site/ Static marketing site deployed to Vercel
build/ App icons and tracked brand resources
docs/ Product and engineering design notes
Contributing
Bug reports and focused pull requests are welcome. Read CONTRIBUTING.md before starting. Please report security issues according to SECURITY.md, not through a public issue.
License
Pin It is available under the MIT License.
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.