Overview
dotagents is a repository that provides a cross-harness convention for agent skills. It contains standard agent skills, configs, harness specs, and setup instructions. The skills are stored in ~/.agents/skills/ and most agent harnesses read this directly without an install step. The repository includes a sync-skills.sh script that symlinks agents, commands, and skills into Claude Code. Most agent harnesses inherit ~/.agents/ natively. It is compatible with Claude Code, Codex CLI, Cursor, copilot, kimi, deepagents, goose, and most other agent harnesses. The repository also provides per-harness specs detailing config, plugins, CLI dependencies, hooks, and permission rules. It includes an agent-mail skill for file-based messaging between agents, and status line scripts for monitoring session details.
Pricing
Free
Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.
Visit WebsitePrices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.
Use Cases
- Teams use dotagents to share a single skillset across multiple agent harnesses running concurrently.
- Teams use dotagents to enable file-based messaging between agents working on separate repositories.
- Teams use dotagents to monitor session status with a status line that displays model, context, tokens, and cost.
- Teams use dotagents to set up Claude Code by symlinking skills, subagents, and commands.
- Teams use dotagents to build new skills using parallel research and spec-first development.
Features
Access reusable AI agent skills
Use agent command examples
Browse agent design examples
Start from provided templates
Use resources with Claude Code
Use resources with agent harnesses
Reviews
No reviews yet. Be the first to share your experience!