edit-guides-surveys
Pass
Audited by Gen Agent Trust Hub on Nov 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes content for guides and surveys retrieved from external API calls. This creates a surface where instructions could theoretically be embedded in the content being managed.
- Ingestion points: Data is ingested through the guides/get, surveys/get, guides/list, and surveys/list commands via the run_amp tool in SKILL.md.
- Boundary markers: The instructions do not specify explicit delimiters for the content being read, but they rely on a structured JSON schema for all interactions.
- Capability inventory: The run_amp tool has the capability to create, update, and archive content via the Amplitude Developer API as documented in SKILL.md.
- Sanitization: The instructions explicitly state that inline HTML in markdown blocks is sanitized, which mitigates the risk of script injection in the managed content.
Audit Metadata