research-subagent

Pass

Audited by Gen Agent Trust Hub on Mar 15, 2026

Risk Level: SAFE
Full Analysis
  • [PROMPT_INJECTION]: The skill implements a web research workflow that ingests untrusted data from external websites using web_fetch and mcp__playwright__snapshot. This constitutes an indirect prompt injection surface.
  • Ingestion points: Web content fetched from external URLs via search and browser tools as described in SKILL.md.
  • Boundary markers: Absent; the instructions do not specify the use of delimiters or explicit 'ignore' instructions for the fetched content.
  • Capability inventory: The agent has the ability to perform further searches and navigate to new URLs based on the content it analyzes.
  • Sanitization: No sanitization or validation of the retrieved web content is mentioned.
  • Note: This behavior is consistent with the primary purpose of a research agent, and no active malicious instructions were found.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 15, 2026, 01:16 PM