agent-skills
Otheraddyosmani/agent-skills
Workflows and best practices for AI coding agents.
Overview
A collection of structured skills that encode workflows, quality gates, and best practices for AI coding agents. Covers the entire development lifecycle from definition to shipping.
README Preview
# Agent Skills\n\n**Production-grade engineering skills for AI coding agents.**\n\nSkills encode the workflows, quality gates, and best practices that senior engineers use when building software. These ones are packaged so AI agents follow them consistently across every phase of development.\n\n```\n DEFINE PLAN BUILD VERIFY REVIEW SHIP\n ┌──────┐ ┌──────┐ ┌──────┐ ┌──────┐ ┌──────┐ ┌──────┐\n │ Idea │ ───▶ │ Spec │ ───▶ │ Code │ ───▶ │ Test │ ───▶ │ QA │ ───▶ │ Go │\n │Refine│ │ PRD │ │ Impl │ │Debug │ │ Gate │ │ Live │\n └──────┘ └──────┘ └──────┘ └──────┘ └──────┘ └──────┘\n /spec /plan /build /test /review /ship\n```\n\n---\n\n## Commands\n\n7 slash commands that map to the development lifecycle. Each one activates the right skills automatically.\n\n| What you're doing | Command | Key principle |\n|-------------------|---------|---------------|\n| Define what to build | `/spec` | Spec before code |\n| Plan how to build it | `/plan` | Small, atomic tasks |\n| Build incrementally | `/build` | One slice at a time |\n| Prove it works | `/test` | Tests are proof |\n| Review before merge | `/review` | Improve code health |\n| Simplify the code | `/code-simplify` | Clarity over cleverness |\n| Ship to production | `/ship` | Faster is safer |\n\nSkills also activate automatically based on what you're doing — designing an API triggers `api-and-interface-design`, building UI triggers `frontend-ui-engineering`, and so on.\n\n---\n\n## Quick Start\n\n\nClaude Code (recommended)\n\n**Marketplace install:**\n\n```\n/plugin marketplace add addyosmani/agent-skills\n/plugin install agent-skills@addy-agent-skills\n```\n\n> **SSH errors?** The marketplace clones repos via SSH. If you don't have SSH keys set up on GitHub, either [add your SSH key](https://docs.github.com/en/authentication/connecting-to-github-with-ssh/
FAQ (2)
how_toHow can I automatically load AGENTS.md in OpenCode similar to Claude Code's session_start hook?
Use the OpenCode plugin hook experimental.chat.messages.transform to inject the content of AGENTS.md into the chat. Create a plugin that reads the AGENTS.md file and prepends it to the message list. See the superpowers plugin for reference: [superpowers.js#L120](https://github.com/obra/superpowers/blob/f2cbfbefebbfef77321e4c9abc9e949826bea9d7/.opencode/plugins/superpowers.js#L120). Example: api.on('experimental.chat.messages.transform', async (messages) => { const content = await fs.readFile('AGENTS.md', 'utf-8'); messages.unshift({ role: 'user', content }); });
TroubleshootingWhy does Claude Code say 'Unknown skill' for all skills in the agent-skills plugin?
The plugin's plugin.json does not explicitly declare the skills and agents directories, which may prevent Claude Code's plugin loader from discovering them. To fix, edit .claude-plugin/plugin.json and add "skills": "./skills" and "agents": "./agents" alongside the existing "commands" field. Save the file and restart the Claude Code session. This workaround is pending confirmation but mirrors how other plugins successfully expose skills.