Agent Skills

Engineering skills for coding agents

The lifecycle,
one slice at a time.

Nine commands that map to how the work actually moves: define it, plan it, build it, prove it, review it, then ship.

Install into Claude Code, Cursor, Codex, Copilot and other agents through the open skills CLI.

npx skills add addyosmani/agent-skills Browse the skills

Command and principle

As published by the project
/spec

Define what to build

Spec before code.

/plan

Plan how to build it

Small, atomic tasks.

/build

Build incrementally

One slice at a time.

/test

Prove it works

Tests are proof.

/review

Review before merge

Improve code health.

/ship

Ship to production

Faster is safer.

Six of the nine commands are shown. Skills also activate on their own based on the work in front of them.

About build auto

The verification stays. Running the plan in one approved pass removes the human stepping between tasks, not the gates: each task is still test driven and committed individually, and it pauses on failures or risky steps.