Addy Osmani's Agent-Skills: Production-Grade Skills for AI Coding Agents

Addy Osmani's Agent-Skills: Production-Grade Skills for AI Coding Agents

Quick Answer: Agent-Skills is an open-source project by Addy Osmani (Google Chrome DevRel) that provides production-grade engineering skills for AI coding agents. It focuses on encapsulating quality gates, best practices, and structured workflows into reusable skill files. The project uses the SKILL.md format and is compatible with Claude Code, Codex CLI, and other SKILL.md agents.

---

Addy Osmani, one of the most recognized names in web development (Google Chrome engineering lead, author of Learning JavaScript Design Patterns), just published "agent-skills" on GitHub. It went trending within days.

The project tackles a specific gap: most AI agent skills are written by individuals for their own workflows. They work, but they don't encode the kind of engineering rigor you'd expect from a production codebase. Agent-Skills aims to change that.

What Agent-Skills does differently

The core idea is encapsulating quality gates into the AI agent's workflow. Instead of writing a skill that says "review this code," Agent-Skills provides skills that build specific checkpoints into each step:

Structured verification. Each skill includes explicit checks the agent must complete before moving on. A code review skill doesn't just flag issues. It verifies the review covered security, performance, accessibility, and maintainability before producing output.

Production-grade patterns. The skills encode patterns from real production environments, not tutorial-level examples. Error handling, edge cases, rollback procedures, and failure modes are built into the instructions.

Workflow encapsulation. Complex multi-step procedures (deployment, migration, incident response) are packaged as single skills that the agent follows step by step, with decision points and branching logic.

Skills to install right now

code-reviewer

Free

Run a structured code review on your recent changes without waiting for a teammate. This skill checks for security vulnerabilities (SQL injection, XSS, authentication bypasses), logic errors, edge cases, performance issues, and style violations.Findings are organized by severity: Critical, Warning, and Suggestion. Each finding includes the file, line number, a description of the issue, and a concrete fix. Use it as a first pass before peer review, or as your only reviewer on solo projects.

Get this skill

git-commit-writer

Free

Stop writing vague commit messages. This skill reads your actual staged diff and generates precise, informative commit messages following the Conventional Commits specification. It detects the commit type (feat, fix, refactor, docs, chore, etc.), identifies the scope from the changed files, flags breaking changes, and suggests splitting commits when multiple logical changes are staged. Works with any git repository.`

Get this skill

readme-generator

Free

Point this skill at any project and it generates a real README from your actual codebase. It scans package.json, pyproject.toml, Cargo.toml, or equivalent to detect your language, framework, and dependencies. It reads your .env.example for setup instructions, checks for Dockerfiles and CI config, and produces a README with only the sections that apply to your project. No placeholder text, no generic filler. Every line is derived from what it actually finds in your repo.

Get this skill

Browse all skills

How it relates to SKILL.md

Agent-Skills uses the SKILL.md format. The same files work in Claude Code, Codex CLI, OpenClaw, Cursor, and every other compatible agent. This isn't a proprietary format or a new standard. It's high-quality content built on the existing open standard.

This matters because the SKILL.md ecosystem has a quality problem. With 13,000+ skills on ClawHub and thousands more scattered across GitHub, finding production-grade skills requires wading through abandoned experiments and half-finished prototypes. Agent-Skills is a curated set from someone with deep expertise in software engineering at scale.

What skills are included

The project focuses on areas where production engineering matters most:

Code quality gates that ensure AI-generated code meets specific standards before being accepted. Not just "does it compile" but "does it handle errors, respect boundaries, and follow the project's patterns."

Deployment workflows that include pre-deploy checks, staged rollouts, and automated rollback conditions. The agent follows a production deployment checklist, not a tutorial deployment.

Testing strategies that go beyond "write a test for this function" to include integration testing, boundary testing, and regression prevention.

Incident response procedures that the agent can follow during production issues, including diagnosis steps, communication templates, and postmortem frameworks.

Why this matters for the SKILL.md ecosystem

Addy Osmani publishing production-grade skills validates two things:

First, the SKILL.md format is mature enough for serious engineering use. This isn't a toy format. It's being used by one of the most respected engineers in the industry to encode real production workflows.

Second, quality differentiation is becoming the defining factor in the skills ecosystem. The gap between a casual skill and a production-grade skill is enormous, and developers are starting to recognize that difference.

For marketplace operators like Agensi, this raises the bar on what "curated" means. Security scanning catches dangerous skills. Quality curation needs to surface the best ones.

How to use Agent-Skills

Clone the repository and copy the skills you need into your agent's skills directory:

git clone https://github.com/nicholasgsmith/agent-skills.git
cp -r agent-skills/skills/code-quality ~/.claude/skills/

Or install individual skills from Agensi where compatible production-grade skills are security-scanned and installable with a single curl command:

mkdir -p ~/.claude/skills && curl -sL https://www.agensi.io/api/install/<slug> | tar xz -C ~/.claude/skills/

What to watch next

Agent-Skills is early. The initial set covers foundational engineering workflows. As the project grows, expect more specialized skills for specific frameworks, languages, and infrastructure patterns.

The broader trend is clear: AI agent skills are moving from "cool hack" to "engineering tool." The projects that take quality seriously will define how developers work with agents over the next few years.

---

Browse 300+ security-scanned SKILL.md skills, including production-grade engineering skills, at agensi.io/skills.