Get started
Use skills with your coding agent
Give your agent one instruction:
Fetch https://skills.internetcomputer.org/llms.txt and follow its instructions when building on ICPThe agent starts using skills right away by fetching them on demand — nothing for you to install. It then offers to set up how the project keeps using them going forward — auto-update, pin, or keep fetching on-demand — and runs whatever the chosen option needs itself (for examplenpx skills add when you pin). It works in any project, including one that already has anAGENTS.md: it appends its own section and never edits yours.
Building tooling instead? Parse the catalogue programmatically:
GET https://skills.internetcomputer.org/api/skills.jsonStart a new project with icp-cli
Projects scaffolded withicp-cliship an AGENTS.md (and CLAUDE.md) that configures itself on first use: your agent asks how you want to use ICP skills — auto-update every session, pin version-locked, or fetch on-demand — and records the choice so every later session just follows it.
See how the modes work and why.
Pin skills into an existing project
Version-lock skills into a repo you already have. Auto-detects your agent (Claude Code, Cursor, Copilot, and more) and writes the skills into its rules files:
npx skills add dfinity/icskillsOr install a single skill by name:
npx skills add dfinity/icskills --skill ckbtcBrowse and read skills yourself
- Browse the skills index and find something that matches your task.
- Open the skill page (for example, ICP CLI).
- Paste the raw Markdown URL into your agent, or share the page URL. Any competent LLM-aware client can fetch either.
Fetching a skill by name
URLs are stable and predictable:
https://skills.internetcomputer.org/skills/{name}/: human-readable HTML pagehttps://skills.internetcomputer.org/skills/{name}/SKILL.md: raw Markdown (Content-Type: text/markdown)https://skills.internetcomputer.org/api/skills/{name}.json: structured metadata (JSON)
Do not construct these URLs by guessing slugs. Fetch /api/skills.jsonor /llms.txt and use the URLs it returns.
Staying current
Subscribe to /feed.xml (RSS) for updates, or watch theupstream repository directly.