skills/cookjohn/wos-skills/wos-export/Gen Agent Trust Hub

wos-export

Pass

Audited by Gen Agent Trust Hub on Sep 30, 2026

Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill invokes a local Python script (push_to_zotero.py) provided within the skill package to facilitate data transfer to Zotero. It utilizes shell piping (echo | python) to pass data, which is a common local integration pattern.
  • [INDIRECT_PROMPT_INJECTION]:
  • Ingestion points: The skill processes paper metadata (titles, abstracts, keywords) sourced from external Web of Science search results via previous tool outputs.
  • Boundary markers: Paper data is handled as a structured JSON object, though there are no explicit delimiters to isolate the text fields from potentially containing malicious instructions.
  • Capability inventory: The skill is capable of executing local Python scripts and making network requests to the local Zotero API (127.0.0.1:23119).
  • Sanitization: The implementation relies on standard json library parsing which provides structural validation but does not filter content within the metadata strings.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 30, 2026, 01:43 AM