Skip to main content
Fingerprint offers three AI tools that work with the coding agents and assistants you already use. Skills and the CLI help you add Fingerprint to your app, and the MCP server connects your agent to your Fingerprint data.

Skills

Give your coding agent specialized knowledge about Fingerprint.

CLI

Integrate and configure Fingerprint from your terminal.

MCP server

Connect your AI assistant to your Fingerprint workspace.

​
Add Fingerprint with a prompt

The fastest way to start is to ask your coding agent to integrate Fingerprint for you. Copy this prompt into Claude Code, Codex, Cursor, or another coding agent from your project’s root directory:

Install Fingerprint in your app with an AI coding agent.

Open in Cursor
The agent runs the Fingerprint CLI, which signs you in, provisions API keys, adds frontend identification and server-side verification to your app, and offers to set up a custom subdomain to protect against ad blockers. The agent then offers to connect the MCP server so you can query your events. Review the changes before you commit them.

​
Skills

Fingerprint skills are installable packages that give AI coding agents specialized knowledge about Fingerprint. Once installed, your agent can help you add Fingerprint to your stack, use Smart Signals, serve the JavaScript agent from your own domain, configure the Rules Engine, and more. Skills point your agent to Fingerprint’s live documentation, so it uses the current API.

​
CLI

The Fingerprint CLI integrates Fingerprint into your app with one command. It signs you in, detects your stack, provisions API keys, and writes the integration to your code. You can also use it to print your API keys and to create, verify, and delete custom subdomains.

​
MCP server

The Fingerprint MCP server connects AI assistants such as Claude, Cursor, and VS Code to your Fingerprint workspace. Once connected, you can query identification events, investigate suspicious activity, manage environments and API keys, and set up custom subdomains in natural language. For example:
  • “Show me recent events where a bot was detected.”
  • “Set up metrics.yourwebsite.com as a custom subdomain.”
The managed server is hosted by Fingerprint. You connect by signing in to Fingerprint, and the server sets up API access for your workspace automatically, so you don’t need to create or paste API keys.

​
Give your agent the docs

Fingerprint publishes its documentation in formats designed for language models:
  • llms.txt lists the v4 documentation pages with short descriptions.
  • llms-full.txt contains the full text of the v4 documentation in one file.
Add either file to your agent’s context when you want it to answer questions about Fingerprint without installing skills or the MCP server.

​
Example prompts

With skills or the MCP server set up, try asking your agent the prompts below. Investigating events and creating environments require the MCP server, because it connects your agent to your Fingerprint workspace.

​
Integrate Fingerprint

  • “Add Fingerprint to my Next.js app and verify each event on the server.”
  • “Block sign-ups from bots, VPNs, and tampered browsers using Smart Signals.”
  • “Attach the signed-in user’s ID to every Fingerprint event as a linkedId.”
  • “Set up metrics.yourwebsite.com as a custom subdomain and update my app to use it.”

​
Investigate events

  • “Summarize the activity of visitor VISITOR_ID over the last 7 days.”
  • “Find all events linked to order ORDER_ID and tell me if anything looks suspicious.”
  • “Show me bot detections from the last 6 hours, grouped by bot name.”

​
Manage your workspace

  • “Create a staging environment with its own public and secret API keys.”
  • “List my custom subdomains and check which ones aren’t active yet.”

​
Detect AI agents on your website

The tools on this page help you work with Fingerprint. To detect and verify AI agents and assistants that visit your own website, see AI agent detection and AI assistant detection.