deep-research

Pass

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill implements a structured research workflow using well-known MCP tools (Firecrawl and Exa) for web searching and scraping. These tools are used as intended for gathering public information.
  • [SAFE]: The instructions incorporate defensive quality rules, such as mandatory source attribution, cross-referencing, and the separation of facts from inferences. These practices help mitigate risks associated with processing untrusted web content (indirect prompt injection).
  • [SAFE]: No suspicious code execution, persistence mechanisms, or obfuscated patterns were detected. The functionality to save reports to files is consistent with the primary purpose of a research tool.
  • [SAFE]: The skill does not attempt to install external packages or execute remote scripts; it relies on pre-configured MCP tools for its operations.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 24, 2026, 07:43 AM
Security Audit — agent-trust-hub — deep-research