suprasumo-design-refact
Pass
Audited by Gen Agent Trust Hub on Nov 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill reads existing application UI and code to inform its refactoring process, which acts as an ingestion point for untrusted data. \n
- Ingestion points: Reads the current UI structure, routes, and components (SKILL.md, step 1). \n
- Boundary markers: There are no specific delimiters or instructions provided to the agent to ignore embedded commands within the ingested code. \n
- Capability inventory: The skill has the ability to modify existing source code, create new files, and control a browser. \n
- Sanitization: No validation or sanitization of the ingested code structure is described before the refactoring is applied. \n- [COMMAND_EXECUTION]: The skill performs file system modifications and browser automation as part of its core functionality. \n
- Evidence: It modifies the existing application codebase (step 6) and opens a browser to verify the changes (step 7).
Audit Metadata