02 — skills
Model-invoked capabilities you install with one command. Curated skills from the open ecosystem — install with npx skills add across Cursor, Antigravity, Codex CLI, and Claude Code.
code-review
Review agent-generated pull requests — priority triage, 5 red flags (CI gaming, code reuse blindness, hallucinated correctness, agentic ghosting, untrusted input), and security checks.
Add -a cursor or -a claude-code to target a specific agent.
security
Secure GraphQL client code — typed variables over string interpolation, plus auth, input validation, and hardening.
Add -a cursor or -a claude-code to target a specific agent.
i18n
CLDR-based formatting for currency, dates, numbers, units, and addresses across 20+ locales.
Add -a cursor or -a claude-code to target a specific agent.
No skills found matching your search.
Skills are reusable instruction sets the agent decides when to invoke. Slash commands are explicit — you trigger them by name. Use skills when you want ambient, procedural knowledge; use commands when you want control.
01
installs to Cursor, Claude Code, Codex, and 60+ more agents via the skills.sh ecosystem.
02
We pick skills worth knowing about — focused on quality over quantity, with clear install paths and source links.
03
Skills teach agents how to do specialized work: localization, design systems, TDD workflows, and more — loaded when the model decides they apply.