MCP Tools DirectoryAbout this integration
Local OCR & screen understanding for agents: read any window, find what to click. No uploads.
- Transport
- stdio
- Authentication
- none
- Initial setup
- user-authorization
- Runtime
- unattended
- Evidence
- documented
- Version
- 0.7.0
- Package
- macos-vision-mcp
- Last compatibility test
- Not independently tested
Connect your agent
macos-vision-mcpPublisher README and executable manifest reviewed from pinned captures. No service credential is required, but macOS Screen Recording and some tools' Accessibility permissions require user authorization. The package was not executed or independently security-audited.
Capabilities: Extract local text and document structure with Apple Vision, Detect barcodes, faces, and document boundaries, Inspect and assert macOS UI content without returning image bytes
Connected profiles
Additional details
io.github.woladi/macos-vision-mcp
Source ↗ · Checked 2026-09-170.7.0
Source ↗ · Checked 2026-09-17CC0-1.0; package licenses are separate
Source ↗ · Checked 2026-09-17npm
Source ↗ · Checked 2026-09-17macos-vision-mcp
Source ↗ · Checked 2026-09-200.7.0
Source ↗ · Checked 2026-09-20Once configured, the MCP client launches the npx command and exposes the local tools after restart; processing is documented as offline after installation.
Source ↗ · Checked 2026-09-20The publisher explicitly documents no API keys or uploads. macOS permission grants are local user authorization rather than service authentication.
Source ↗ · Checked 2026-09-20The reviewed package is configured as a client-launched local command using npx; no remote MCP endpoint is used for this listing.
Source ↗ · Checked 2026-09-20Publisher documentation reviewed; package and integration endpoint not executed or independently security-audited.
Source ↗ · Checked 2026-09-20