Skip to main content

AI Assist

Drop AddressZen into any AI workflow, from coding agents driving a repo to chat-driven app builders generating a frontend.

Quick start​

  1. Give your agent the docs. Paste llms.txt into the model's context.
  2. Get an API key.
    • Quick test: use ak_test. Public, capped at 5 lookups/day, will break in production.
    • Real testing: sign up for a free trial key with a real test balance.
  3. Install the skills. npx skills add addresszen/skills teaches Claude Code, Cursor and Codex the API, widgets and CLI.
  4. Lock it down before going live. Restrict the key to your origin via Allowed URLs.

Resources​

Ordered by what an agent should reach for first.

  • llms.txt: full API reference, integration guides and widget docs as one file. Paste into a model's context, or wire into a docs MCP server like Context7.
  • Agent skills: npx skills add addresszen/skills. Skills for Claude Code, Cursor and Codex: azn-address-lookup, azn-api-integration, azn-react, azn-cli, azn-js-sdk and azn-python-sdk. The skills index lists them for agents.
  • CLI: npm install -g @addresszen/cli. Built for agents: JSON output in non-TTY, stable exit codes, no prompts under pipes. Subcommands: auth, keys, configs, find, verify, email, phone, doctor. Set AZN_API_KEY (and AZN_USER_TOKEN for key management) rather than running azn auth login. Use a backend key (no Allowed URLs).
  • MCP server: coming soon. Will surface API docs, dataset metadata and key management as MCP tools.
  • Address Lookup: international address autocomplete widget.
  • Per-page markdown: append .md to any docs URL for the plain-text version (e.g. /docs/address-lookup.md).

AI builder guides​

Per-platform setup, prompts and key-storage instructions live in Build with AI integrations.

Coding Agents​

AI App Builders​

Need help?​