# usezn Developer Portal

usezn protects AI agent tool calls from prompt injection before they reach your tools.

## Start

1. Create an API key at https://usezn.com/dashboard/
2. Store it as `ZN_API_KEY`.
3. POST untrusted tool-call input to `https://api.usezn.com/v30/analyze`.
4. Stop execution when `verdict` is `block`.

```bash
curl -X POST https://api.usezn.com/v30/analyze \
  -H "Authorization: Bearer $ZN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"input":"text from an untrusted source"}'
```

Default endpoint applies rules plus the English neural model; swap v30 for prod in the URL for pure deterministic rules.

## MCP

The Streamable HTTP endpoint is `https://api.usezn.com/mcp`. Use JSON-RPC POST with `tools/list` or `tools/call`; authenticate with the same Bearer API key. Manifest: https://usezn.com/.well-known/mcp.json

## Resources

- Docs: https://usezn.com/docs/
- OpenAPI: https://usezn.com/openapi.json
- Agent guide: https://usezn.com/agents.md
- Sandbox: https://usezn.com/playground/
- CLI: [zn-gate on npm](https://www.npmjs.com/package/zn-gate), install with `npm install -g zn-gate`
- Contact: https://usezn.com/contact/