Back to directory

AgentSync

A cross-platform CLI written in Rust that keeps one set of AI coding-agent instructions, commands and skills in a single directory and symlinks them out to the different locations that tools such as Claude Code, GitHub Copilot, Cursor and Codex each expect.

AgentSync

AgentSync maintains a single source of truth in a .agents directory and creates symbolic links from it to the various file paths that different AI coding tools read, such as CLAUDE.md and .claude/skills for Claude Code, .gemini/commands for Gemini CLI, or .cursor/rules for Cursor, so a change in one place propagates to every tool instantly rather than needing to be copied to each one by hand. Configuration is stored in TOML, and the project's own .gitignore is updated automatically as files are added.

It ships as a single static binary with no runtime dependency, backs up existing files before replacing them, and skips gracefully in a continuous-integration environment where the binary is unavailable. It can also install curated skills from a skills collection or an external provider, and includes an end-to-end catalog check that every skill entry can still be resolved, installed and registered correctly.

It is written in Rust, distributed for Linux, macOS and Windows, and licensed under MIT.

More in Agent skills

See all 581
  • Provenance gate

    github.com

    A deterministic authorization gateway that sits between an AI agent and its tools, tracing where each parameter of a side-effect call came from, and blocking the call or asking a human when the source does not check out.

  • A markdown framework of system rules that restates Isaac Asimov's Four Laws of Robotics as a set of prioritized loss calculations for large-language-model agents to follow, published as a discovery file for agents and crawlers to read.

  • Vitruvius

    github.com

    An engineering research agent that runs a discover, read, synthesize, verify and review loop over technical questions, reading sources directly rather than inferring from titles, and publishing its own scored verifier benchmark, weaknesses included.