ux-flow
Pass
Audited by Gen Agent Trust Hub on Aug 10, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it processes untrusted user input to influence the generation of project files and design artifacts.
- Ingestion points: Untrusted data enters the agent context through the
$0and$ARGUMENTSvariables inSKILL.md, which capture user-provided flow names and descriptions. - Boundary markers: The instructions do not define clear delimiters or use explicit 'ignore' directives to separate user-provided data from the agent's core instructions, increasing the risk that embedded instructions in the user input could be obeyed.
- Capability inventory: The skill directs the agent to generate ASCII diagrams and actual page files, granting it write-access capabilities to the project directory based on the processed input.
- Sanitization: No input validation, filtering, or escaping mechanisms are specified to ensure user input is treated as passive data rather than active instructions.
Audit Metadata