Skill files (.claude/skills/, .cursor/rules/*.mdc, AGENTS.md, .windsurfrules) are becoming a core part of AI-assisted development workflows. The problem: they scatter across projects, diverge silently, and every new repo means rebuilding behavioral config from scratch. Each tool uses a different format and location.

Skilldeck keeps one local library and deploys to any tool in the correct format automatically. Ten built-in target profiles cover Claude Code, Cursor, Copilot, Windsurf, Codex, and more. Drift detection shows when a deployed skill has fallen out of sync with the library version. Bidirectional sync lets you pull improvements back from a project into the library.

The interesting part technically: the entire app was built by Claude Code using a harness engineering methodology — a ground truth JSON file, Playwright E2E verification tests, a regression gate based on a surfaces map, and a feature intake protocol. 31 features across multiple autonomous sessions with no manual application code. I wrote two articles about the harness approach if that side is interesting.

No cloud, no backend, local filesystem only. Windows/macOS/Linux. Open source.

github.com/ali-erfan-dev/skilldeck