ui-visual-tester
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONNO_CODE
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes external project documentation, requirements, and architecture to design test cases. This ingestion of untrusted data represents a potential attack surface.\n
- Ingestion points: Rule 1 in SKILL.md mandates reading project-level instructions, architecture, and requirements.\n
- Boundary markers: There are no instructions for using delimiters or boundary markers for external project data.\n
- Capability inventory: The skill instructs the agent to use authorized tools for non-destructive checks; no scripts or executable code are bundled with the skill.\n
- Sanitization: No explicit sanitization or filtering of input documentation is performed.\n- [NO_CODE]: The skill consists entirely of markdown and CSV templates and does not contain any executable scripts or configuration files that could run code.\n- [SAFE]: The skill contains explicit guidelines to prevent unauthorized destructive actions, credential exposure, and the fabrication of test evidence.
Audit Metadata