yutori-scout
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to gather user inputs and external documentation to construct monitoring queries, which constitutes a potential injection surface.\n
- Ingestion points: User-provided monitoring goals and external documentation or reference pages (SKILL.md).\n
- Boundary markers: Absent; there are no instructions to use delimiters for untrusted data.\n
- Capability inventory: Uses the
create_scouttool to register scheduled web-monitoring tasks.\n - Sanitization: No sanitization or validation of the external content is mentioned.\n- [DATA_EXPOSURE_AND_EXFILTRATION]: The skill facilitates configuring webhooks for data updates. It includes a security best-practice instruction requiring the agent to confirm the destination URL and ownership with the user before setup.\n- [EXTERNAL_DOWNLOADS]: The skill suggests fetching documentation from Cloudflare-hosted sites, which is a well-known and reputable service provider.
Audit Metadata