simplify
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content, flows, and interfaces to produce simplified versions. \n
- Ingestion points: The skill takes content, flow descriptions, or interface definitions as input in SKILL.md. \n
- Boundary markers: The instructions do not define specific delimiters or warnings to ignore embedded instructions in the source text. \n
- Capability inventory: No capabilities for file writing, network operations, or shell execution were identified in the skill instructions. \n
- Sanitization: The skill does not perform explicit sanitization of the input text. However, without harmful capabilities, the risk of instruction following is negligible.
Audit Metadata