yutori-research

Pass

Audited by Gen Agent Trust Hub on May 2, 2026

Risk Level: SAFE
Full Analysis
  • [DATA_EXFILTRATION]: The tool 'run_research_task' includes a 'webhook_url' parameter to notify external services (e.g., Slack, Zapier) upon task completion. While this involves sending data to non-whitelisted domains, it is a standard functional component for asynchronous research workflows.
  • [PROMPT_INJECTION]: Because the skill is designed to synthesize and present findings from web searches, it inherently processes untrusted data which serves as a surface for indirect prompt injection. No specific evidence of malicious instructions or lack of sanitization was found in the static instructions.
Audit Metadata
Risk Level
SAFE
Analyzed
May 2, 2026, 09:45 PM