deep-research

Pass

Audited by Gen Agent Trust Hub on Mar 11, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The skill uses the bash tool to execute a Python script (run.py) which manages the research workflow.
  • [EXTERNAL_DOWNLOADS]: The skill fetches content from arbitrary external URLs provided in the fetch_urls parameter and retrieves data from the Tavily search API. These operations are essential to its primary purpose as a research tool.
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. Ingestion points: Data enters the agent's context through search results in run.py and external website content fetched via _fetch_page. Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the fetched content. Capability inventory: The skill has network access and operates via bash command execution. Sanitization: The script uses regular expressions to strip HTML tags but does not validate the text content for malicious instructions that could influence the LLM's final summary output.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 11, 2026, 09:19 PM