AI/TLDR

Addy Osmani · 2026-05-03 · major

Addy Osmani's Agent Skills — Senior-Engineering Workflow Scaffolding for AI Coding Agents

Google engineer Addy Osmani publishes a 20-skill framework that forces AI coding agents through phases of spec, plan, build, test, review and ship instead of the shortest path to done.

Addy Osmani Agent Skills cover graphic showing a six-phase agent workflow
addyosmani.com

20 opinionated skills that force coding agents through spec, plan, build, test, review and ship phases.

Key specs

LicenseMIT
GitHub stars28,130
Skills20
Phases6
Hn points207
Hn comments88

What is it?

An MIT-licensed pack of 20 skills and 7 slash commands (/spec, /plan, /build, /test, /review, /code-simplify, /ship) that wraps a coding agent in the workflow scaffolding a senior engineer would use. Drops into Claude Code, Cursor, Gemini CLI, Windsurf and other agents that read the Anthropic skills format.

How does it work?

Each skill is a markdown file with a defined process, an anti-rationalization table that calls out the shortcuts agents try to take, and verification requirements. The slash commands chain skills across six phases — Define, Plan, Build, Verify, Review, Ship — so an agent cannot mark a task done without producing the artefacts (spec doc, test plan, review notes) the phase requires.

Why does it matter?

Coding agents default to taking the shortest path to a passing test, which is how production incidents ship. Forcing the invisible parts of senior engineering work — scope discipline, written specs, test plans, review notes — into the agent's loop is the cheapest way to raise the floor on agent-written code right now.

Who is it for?

Teams shipping agent-written code into real codebases.

Try it

npx skills add addyosmani/agent-skills

Sources · 3 outlets

Tags

  • agent-skills
  • claude-code
  • coding-agents
  • engineering-practices
  • anti-rationalization
  • addy-osmani

← All releases · Learn AI