design-product-experiment
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill provides instructions for the agent to ingest and process potentially untrusted external data, creating a surface for indirect prompt injection.\n
- Ingestion points: The
SKILL.mdfile instructs the agent to use external signals, survey responses, and documents to refine experimental hypotheses.\n - Boundary markers: The instructions do not provide delimiters (such as XML tags or unique markers) or specific directives to ignore instructions embedded within the ingested data.\n
- Capability inventory: The skill itself is instructional markdown and does not include subprocess executions, network operations, or file-writing scripts within its own files.\n
- Sanitization: There are no explicit requirements or steps for the agent to validate or sanitize the content of the processed survey results or documents.
Audit Metadata