content-extract

Pass

Audited by Gen Agent Trust Hub on Mar 9, 2026

Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
  • [PROMPT_INJECTION]: The skill processes untrusted data from external URLs, creating an indirect prompt injection surface.\n
  • Ingestion points: Content is retrieved from remote URLs supplied via the args.url argument in scripts/content_extract.py.\n
  • Boundary markers: The skill does not implement boundary markers or instructions to isolate retrieved content from the agent's logic.\n
  • Capability inventory: The skill can execute subprocesses via scripts/content_extract.py and perform network operations.\n
  • Sanitization: There is no evidence of sanitization or filtering of the extracted Markdown content to remove potentially malicious instructions.\n- [COMMAND_EXECUTION]: The script scripts/content_extract.py executes an external script using a path determined at runtime.\n
  • Evidence: The _find_mineru_wrapper function dynamically computes the path to a sibling skill's script (mineru_parse_documents.py) based on the filesystem layout or environment variables.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 9, 2026, 10:14 PM