deep-research

Pass

Audited by Gen Agent Trust Hub on Apr 6, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill defines a logical and secure workflow for automated research. It utilizes specific tools like firecrawl and exa for web search and scraping, which is consistent with its stated purpose.- [COMMAND_EXECUTION]: The instructions involve the use of MCP tools for web interaction. These operations are scoped to research queries and URL content retrieval, following established patterns for search agents.- [DATA_EXFILTRATION]: No unauthorized data transfer or credential harvesting patterns were detected. The skill's network operations are restricted to searching and fetching content for the user's research report.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 6, 2026, 04:00 AM
Security Audit — agent-trust-hub — deep-research