deep-research

Pass

Audited by Gen Agent Trust Hub on Mar 24, 2026

Risk Level: SAFEPROMPT_INJECTION
Full Analysis
  • [SAFE]: The skill utilizes well-known and expected tools such as Firecrawl and Exa MCPs for its intended purpose of conducting multi-source web research and synthesis.
  • [PROMPT_INJECTION]: The skill exposes a surface for indirect prompt injection as it processes untrusted content from the internet.
  • Ingestion points: Untrusted data enters the agent context through the use of firecrawl_search, firecrawl_scrape, web_search_exa, and crawling_exa tools (SKILL.md).
  • Boundary markers: There are no explicit instructions or delimiters defined to isolate the scraped content or warn the agent against executing embedded commands within that content.
  • Capability inventory: The skill is capable of executing multiple search and scrape tool calls and can launch parallel sub-agents using the Claude Code Task tool (SKILL.md).
  • Sanitization: The workflow does not include specific validation, filtering, or sanitization steps for the data retrieved from external URLs prior to its synthesis into reports.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 24, 2026, 07:41 AM
Security Audit — agent-trust-hub — deep-research