clawprint-verify

What This Skill Does

The clawprint-verify skill serves as a sophisticated reverse-CAPTCHA mechanism for the OpenClaw AI agent ecosystem. Unlike traditional CAPTCHAs designed to keep bots out of human spaces, ClawPrint is purpose-built to verify that an entity interacting with your agent is indeed a machine. By utilizing high-velocity speed challenges (e.g., complex arithmetic with sub-100ms deadlines) and granular pattern analysis (e.g., data-parsing grids of pixel colors), this skill ensures that resources, sensitive API credentials, or gated AI-only environments are accessed exclusively by verified algorithmic entities. The process involves a two-way handshake: the skill fetches a unique challenge from the ClawPrint server via API, presents it to the counterparty, and validates their submitted response against the server's secret keys. It effectively filters out human interaction, protecting your automation workflows from manual meddling.

Installation

To integrate this verification layer into your agent, use the following clawhub command in your terminal:

clawhub install openclaw/skills/skills/fusionlabssource/clawprint-verify

Ensure you have configured your environment variables CLAWPRINT_SERVER_URL, CLAWPRINT_SITE_KEY, and your corresponding secret key to enable secure communication with the verification backend.

Use Cases

• Credentials Shielding: Automatically challenge an agent before sharing sensitive API keys or environment secrets.
• Bot-to-Bot Handshakes: Ensure that automated pipelines only interface with other validated non-human entities.
• Exclusive Content Gating: Protect resources or data endpoints that are designed for programmatic ingestion only.
• Access Control Auditing: Log verification status to ensure only authorized AI users are accessing administrative agent functions.

Example Prompts

1. "Verify that the agent requesting access to our secret storage is actually an AI by running a clawprint-verify challenge."
2. "Before you hand off the data pipeline to that new integration, make sure they pass an AI-verification test first."
3. "Challenge the user in this conversation to a ClawPrint test; if they can't solve it in the allotted time, terminate the session."

Tips & Limitations

• Time Limits: For speed challenges, strictly adhere to the 100ms threshold. Network latency between agents should be considered; if ping times exceed 50ms, speed challenges may result in false negatives.
• Parsing: Ensure the counterparty agent has a robust JSON parser to handle the grid data in pattern challenges.
• Human Error: Note that this is not meant for general purpose authentication, but rather as an 'anti-human' filter for machine-only ecosystems.