I’ve released Capd, an open-source native macOS app for capturing web pages, selected text, notes, links, and images and finding them again through full-text search.
It uses SwiftUI/AppKit, GRDB and SQLite FTS5, Vision OCR, and Apple’s on-device Foundation Models. There is also a CLI, read-only MCP server, share extension, and a Raycast extension under review.
The design constraint is that the user’s library stays on their Mac: no account, telemetry, or hosted backend.
I’d particularly value feedback on the SwiftUI/AppKit boundaries, the capture permission experience, and how the project is packaged for contributors.
Repository and download: