Skills
Navigate
skills/didit-protocol/skills/didit-face-match/Gen Agent Trust Hub

didit-face-match

Pass

Audited by Gen Agent Trust Hub on Sep 14, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill communicates with Didit's official API endpoints at verification.didit.me and apx.didit.me for biometric comparison and account management.
  • [COMMAND_EXECUTION]: The skill includes a Python script (scripts/match_faces.py) that utilizes the requests library to send image data to the Didit API and process the results.
  • [INDIRECT_PROMPT_INJECTION]: The skill processes external image data to generate facial comparison scores and metadata.
  • Ingestion points: Image files provided via command-line arguments in scripts/match_faces.py.
  • Boundary markers: The script uses a strict multipart/form-data schema for API requests and expects a structured JSON response.
  • Capability inventory: HTTP network operations via the requests library in scripts/match_faces.py.
  • Sanitization: The script performs basic file path validation using os.path.isfile before processing images.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 14, 2026, 08:44 PM
Security Audit — agent-trust-hub — didit-face-match