drawing-analyzer

Pass

Audited by Gen Agent Trust Hub on Mar 5, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted external PDF files, creating an indirect prompt injection surface where text inside a drawing could attempt to influence the agent's behavior.
  • Ingestion points: The analyze_pdf_drawing method in SKILL.md ingests text directly from user-provided PDF files.
  • Boundary markers: There are no explicit boundary markers or instructions to ignore extracted text content when presenting data to the agent.
  • Capability inventory: The skill is restricted to file reading and data processing; it does not have capabilities for network operations or subprocess execution.
  • Sanitization: The implementation uses regex for extraction but does not perform sanitization of the extracted strings to prevent prompt-based attacks.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 5, 2026, 04:28 AM