verify-document

Pass

Audited by Gen Agent Trust Hub on Oct 27, 2026

Risk Level: SAFEDATA_EXFILTRATIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
  • [DATA_EXFILTRATION]: The skill instructs the agent to upload local document files to a remote endpoint (https://www.stipple.sh/v1/warrants). The documents targeted for verification, such as payslips, bank statements, and identification documents, contain highly sensitive personal and financial data.
  • [EXTERNAL_DOWNLOADS]: The skill performs network operations to the non-whitelisted domain stipple.sh using curl to interact with an API. This communication is essential for the document verification service but represents external data transmission.
  • [INDIRECT_PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it processes external document data which could contain malicious instructions.
  • Ingestion points: The skill processes user-supplied document files (PDF, PNG, JPEG, BMP, TIFF) from local paths or URLs in SKILL.md.
  • Boundary markers: No specific delimiters or instructions are provided to the agent to ignore or isolate embedded text content within the documents being verified.
  • Capability inventory: The skill has network access via curl to transmit data to the Stipple API.
  • Sanitization: There is no evidence of specific sanitization or filtering of document content before it is processed by the agent.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 27, 2026, 07:51 AM