The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests identity document images/PDFs and returns OCR-extracted text (e.g., full name, address, document numbers). This represents a surface for indirect prompt injection if the processed document contains malicious instructions intended to influence the agent.
Ingestion points: front_image and back_image files processed by the verify_id function in scripts/verify_id.py.
Boundary markers: None present in the response schema to delimit or signpost the extracted OCR text for the agent.
Capability inventory: The skill possesses file system read access for document processing and network communication capabilities with the didit.me domain.
Sanitization: No explicit sanitization or safety filtering is performed on the extracted OCR text before it is returned to the agent's context.

didit-id-document-verification