addyosmani

agent-skills

AI#AI Agents#Software Engineering#Workflow Automation#Prompt Engineering

// summary

Agent Skills provides a structured library of engineering workflows designed to help AI coding agents follow professional best practices throughout the software development lifecycle. These 20 modular skills cover everything from initial specification and planning to testing, security, and production deployment. By enforcing rigorous quality gates and anti-rationalization checks, the project ensures that AI-generated code meets the standards of senior engineering teams.

// technical analysis

Agent Skills provides a structured framework of production-grade engineering workflows designed to guide AI coding agents through the entire software development lifecycle. By encoding best practices—such as spec-driven development, test-driven development, and rigorous code review—the project forces agents to move beyond prototype-level output toward reliable, maintainable code. A core technical trade-off is the emphasis on process over speed, utilizing 'anti-rationalization' tables to prevent agents from skipping critical verification steps, thereby ensuring that quality gates are consistently met.

// key highlights

Provides 20 structured skills covering the full development lifecycle from initial idea refinement to production deployment.

Implements anti-rationalization logic that forces agents to address common excuses for skipping best practices like testing or security reviews.

Features seven slash commands that map directly to development tasks, allowing for automated activation of relevant engineering workflows.

Includes specialized agent personas like a Senior Staff Engineer for code reviews and a Security Auditor for vulnerability assessment.

Integrates proven industry methodologies such as the test pyramid, Hyrum's Law, and trunk-based development directly into agent instructions.

Uses a modular design where supplementary checklists for security, performance, and accessibility are loaded only when necessary to minimize token usage.

// use cases

Enforcing consistent development workflows like TDD and spec-driven development across AI agents

Automating quality gates for code reviews, security hardening, and performance optimization

Integrating professional engineering standards into popular AI coding tools like Claude Code, Cursor, and Gemini

// getting started

To begin, choose your preferred AI agent tool from the provided documentation, such as Claude Code, Cursor, or Gemini CLI. You can install the skills via a marketplace plugin, by cloning the repository and pointing your tool to the local directory, or by manually copying the Markdown skill files into your agent's rules configuration. Once installed, you can trigger specific workflows using the provided slash commands or by referencing the skills directly in your agent's context.