capturist-mcp
This server does not currently have a deployable release on Glama. Follow the repository's setup instructions to install and configure it for your MCP client, including any required credentials and local dependencies.
capturist-mcp
Give Claude, ChatGPT, Codex and other AI apps your Mac screenshots. capturist is the command line tool and local MCP server for Capturist, the Mac app that keeps your screenshots, saved text and clipboard in one private library and reads the words inside every screenshot on your Mac.
Ask your AI app things like:
“Fix the failing test from my last screenshot.”
“Find the screenshot with the Acme invoice and read me the total.”
“What did the error I captured an hour ago say?”
It's a local server: everything stays on your Mac. macOS 14 Sonoma or later, Apple silicon or Intel. You need Capturist installed: the free beta, Capturist Complete, or Capturist from the Mac App Store.
Install
curl -fsSL https://capturist.nodetools.org/install-cli.sh | bashThis installs a universal, Developer ID signed and Apple-notarized capturist to ~/.local/bin after checking its checksum and signature. Run it again to update. The same zip is attached to each GitHub release.
Or with Homebrew:
brew install vortex-303/tap/capturistClaude desktop, one click: download capturist.mcpb from the latest release and open it (or Claude › Settings › Extensions › Install Extension).
Listed in the official MCP Registry as io.github.vortex-303/capturist-mcp.
Or build it yourself (Swift 6):
swift build -c release --product capturist
cp .build/release/capturist ~/.local/bin/Related MCP server: ContextPulse
Connect your AI apps
capturist setup claude-desktop # Claude (desktop app)
capturist setup chatgpt # ChatGPT (desktop app) and Codex
capturist setup claude-code # Claude Code
capturist setup codex # Codex CLIThen reopen the app. For any other MCP client:
{
"mcpServers": {
"capturist": { "command": "capturist", "args": ["mcp"] }
}
}Tools
Tool | What it does |
| Newest captures, optionally since a time or of one kind |
| Search names, source app, window title and the text inside screenshots |
| One capture: details, the text inside it, and the image |
| Just the text inside a capture (errors, logs, code) |
| Recent clipboard text, if clipboard history is on in Capturist |
| Add text or an image to the library, only if you allow it in Capturist |
How it works
capturist mcp is a standard MCP server over stdio.
Relay: when Capturist 1.1 or later is open with Settings › AI Agents on, it forwards every message to the app's own server on
127.0.0.1, so your AI apps get every tool, under the app's rules.Direct: when the app is closed, or on Capturist 1.0, it opens your library read-only and reads it itself.
Text inside screenshots is recognized with Apple's on-device Vision framework. See SECURITY.md for the details of the local connection.
Command line
capturist recent --since 30m # newest captures
capturist search "invoice 502" --json # names, apps, window titles, text inside images
capturist text latest # the text inside the latest capture
capturist clips --since 1h # clipboard history (if turned on)
capturist setup shortcut # ⌥⌘7 saves selected text to Capturist
capturist doctor # check access, index and setupFull reference: capturist.nodetools.org/cli. Exit codes: 0 ok · 1 nothing found · 2 bad input · 3 library missing or not readable.
Packages
Module | What it is |
| The command line tool and MCP server |
| MCP tool handler shared by the CLI and the Capturist app |
| Reads (and, in the app, writes) the Capturist library and its text index |
| Shared models |
swift testLicense
MIT © Vortex303. The Capturist app itself is not open source.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
This server cannot be deployed
Maintenance
Related MCP Connectors
Mac & Windows: let ChatGPT, Claude & Cursor use your email, calendar, iMessage, Teams, files.
Memory for your AI: what you saw on screen, said in calls and worked on, searchable by agents.
Search what you've seen on screen and heard in meetings, recorded on your own Mac or PC.
Search photos and video moments by visual meaning, people, dates, and folders across local libraries, Apple Photos, and OneDrive. Manage collections, groups, favorites, people, and duplicate review. Synced metadata remains searchable while the desktop is offline; local and Apple Photos media access needs the owning device online with remote access enabled. OneDrive media can be accessed directly from the cloud. OAuth or Bearer API keys. Paid Cloud Access required; hosted visual search uses prepaid credits. Text-in-image search is semantic matching, not guaranteed exact OCR. Setup: https://organizewith.ai/en/guides/connect-ai-assistants
Related MCP Servers
- AlicenseAqualityAmaintenanceA macOS utility that captures screenshots and analyzes them with AI vision, enabling AI assistants to see and interpret what's on your screen.32,020 npm5,494MIT

ContextPulseofficial
AlicenseAqualityBmaintenanceLets AI assistants understand what you're working on — current screen content, recent dictation, clipboard, and saved notes — running entirely on your own machine with nothing sent to the cloud.362AGPL 3.0- AlicenseNot gradedqualityBmaintenanceEnables AI assistants to search and retrieve memories from your Mac, including screen captures, meeting transcripts, and browsing history, all locally and privately.1Elastic 2.0
- AlicenseNot gradedqualityCmaintenanceEnables AI assistants to perceive and operate macOS applications through screenshots, on-device OCR, accessibility elements, and consent-gated input, and to run an app in an isolated virtual-display session with screenshots, OCR/ASCII frames, bounded history, and evidence reports. It supports text-only models via spatial text representations while keeping capture read-only by default and gating control behind macOS permissions and operator authorization.MIT