didit-face-search

What This Skill Does

The didit-face-search skill integrates the Didit Face Search standalone API into the OpenClaw ecosystem, enabling high-performance 1:N facial biometric verification. This tool compares a provided user image against an existing database of previously verified sessions to identify potential matches. By leveraging advanced facial recognition algorithms, it returns detailed similarity scores and metadata, allowing developers and administrators to implement robust identity management protocols. The tool is specifically designed to handle tasks such as checking for existing accounts, preventing fraudulent multiple registrations, and identifying blocklisted individuals with precision.

Installation

To add this capability to your OpenClaw agent, execute the following command in your terminal or command-line interface within the agent's environment:

clawhub install openclaw/skills/skills/rosasalberto/didit-face-search

Ensure that you have your Didit API key ready from the Didit Business Console, as it is required for authentication during the configuration process.

Use Cases

This skill is indispensable for businesses requiring secure user onboarding and fraud prevention. Use it to:

• Detect duplicate account creations during registration workflows.
• Validate if a user has already undergone identity verification in a previous session.
• Cross-reference new sign-ups against an internal blocklist to automatically trigger account declines.
• Implement biometric deduplication to ensure the integrity of user identity records across your application.
• Facilitate manual review by providing similarity scores and visual match evidence for borderline cases.

Example Prompts

1. "Check if the attached photo of the new registrant is already present in our verified user database to prevent duplicate accounts."
2. "Search for the provided face image against our system and tell me if it matches any blocklisted users or existing sessions with a similarity above 80%."
3. "Scan this image and return the ranked matches from our verification history, including similarity percentages and the associated user details."

Tips & Limitations

• Similarity Thresholds: Treat scores above 90% as high-confidence matches. Use the 70-89% range as a trigger for manual human verification to avoid false positives.
• Image Quality: Ensure uploaded images (JPEG, PNG, WebP, TIFF) are under 5MB and provide a clear, upright frontal view of the face. For non-upright images, utilize the rotate_image parameter.
• Security: Always protect your API keys using environment variables rather than hardcoding them in your scripts.
• Data Privacy: The service handles sensitive biometric data; ensure your usage complies with local privacy laws and your organization's data protection policies.