engineering / skill

Agent Skills

Guide your AI coding agent through spec, plan, build, test, review and ship stages with 25 engineering skills.

At a glance

Purpose
Guide your AI coding agent through spec, plan, build, test, review and ship stages with 25 engineering skills.
Runs in
Local agent with shell
Operating systems
macOS, Windows
Needs
Runtime: Not stated; Needs internet: Not stated; Needs accounts or API keys beyond your agent: Not stated
Files and data
Reads files: Yes; Writes files: Yes; Sends data to other services: Not stated
Tool type
Skill
Delivery
From GitHub
Version
Latest on the main branch
Access
Code on GitHub
Compatibility
According to the README, the skills install into Claude Code, Cursor, Codex, GitHub Copilot, Cline and more through the open skills CLI, which it says reaches 70+ agents. It also gives steps for Antigravity CLI, Gemini CLI, Windsurf, OpenCode, Kiro and Command Code. The README says the skills are plain Markdown and work with any agent that accepts system prompts or instruction files.
External costs
The repository is MIT-licensed and states no pricing. Your coding agent or model provider may charge for usage.
Permissions
The skills act through your coding agent with the access you give it. According to the README, the repository also carries lifecycle hooks for Claude Code and four agent personas for reviews. It says /build auto has you approve the plan once and then runs every task autonomously, committing each task and pausing on failures or risky steps. It also describes a browser skill that uses Chrome DevTools MCP for DOM inspection, console logs, network traces and performance profiling.
Usage terms
MIT License, as stated in the repository. See its LICENSE file for the exact terms.

What it does

Agent Skills is a collection of 25 skills for AI coding agents. According to the README, the skills encode the workflows, quality gates and practices that senior engineers use, and they cover six stages: define, plan, build, verify, review and ship.

The README groups the skills by stage. Examples include spec-driven-development, planning-and-task-breakdown, incremental-implementation, test-driven-development, code-review-and-quality, security-and-hardening, ci-cd-and-automation and shipping-and-launch. A meta skill, using-agent-skills, maps incoming work to the right skill. The README also lists nine slash commands such as /spec, /plan, /build, /test, /review and /ship, four agent personas (code reviewer, test engineer, security auditor and web performance auditor) and seven reference checklists.

The README says every skill follows the same structure: an overview, when to use it, a step-by-step process, a table of common excuses with rebuttals, red flags and evidence requirements.

Who it is for

Developers who use AI coding agents and want them to follow a consistent spec, plan, build, review and ship workflow.

Example tasks

Write a spec before any code with the spec-driven-development skill.
Break a spec into small tasks with acceptance criteria using the planning-and-task-breakdown skill.
Review a change before merging with the code-review-and-quality skill.
Work through failing tests or broken builds with the debugging-and-error-recovery skill.
Prepare a production release with the shipping-and-launch skill.

Getting started

According to the README, one route for any agent is the skills CLI: npx skills add addyosmani/agent-skills installs all 25 skills, and adding --list lets you browse first. For Claude Code, run /plugin marketplace add addyosmani/agent-skills and then /plugin install agent-skills@addy-agent-skills. For Codex, run codex plugin marketplace add addyosmani/agent-skills and then codex plugin add agent-skills@agent-skills. The README gives separate steps for Cursor, Antigravity CLI, Gemini CLI, Windsurf, OpenCode, GitHub Copilot, Kiro and Command Code. The README has the exact commands for the Claude Code, Codex, Antigravity CLI, Gemini CLI and Command Code routes and points to setup pages for the others.

Limitations

The skills need an AI coding agent that can load them. According to the README, the Codex plugin route needs Codex CLI v0.122 or later, and the Claude Code marketplace route clones over SSH unless you use the HTTPS form the README describes. The README says the lifecycle slash commands are Claude Code commands and do not appear in GitHub Copilot, where you invoke skills by name. It also says a per-skill install copies only that skill's folder, so paths to the shared reference checklists are unavailable, and that in affected Antigravity CLI releases the wrapper commands are not discoverable. The repository can change at any time; check its README before you install.

Data practices

The README does not describe data handling. What is sent where depends on your coding agent and model provider. The README mentions an optional, user-authorized cross-model escalation in the doubt-driven-development skill without describing where data goes.

Community experience

Not rated yet

Sign in to share your experience.

Comments 0

No comments yet.

Sign in to comment

Report a listing concern