PDF OCR Extraction

Pass

Audited by Gen Agent Trust Hub on Mar 10, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is designed to ingest and process external PDF files, which introduces a risk of indirect prompt injection. Instructions embedded in a document could be interpreted as commands after OCR processing. * Ingestion points: Untrusted PDF files provided for OCR. * Boundary markers: No delimiters or boundary instructions are defined in the documentation. * Capability inventory: The skill interacts with the office-mcp server for file processing. * Sanitization: No mechanisms for sanitizing or validating extracted text are specified. * Mitigation: Wrap extracted text in clear delimiters and instruct the agent to disregard any embedded instructions within that content.
  • [NO_CODE]: The skill consists entirely of markdown documentation and YAML metadata. No scripts, binaries, or other executable code elements are included in the package.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 10, 2026, 05:18 AM