ui-review
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill involves running the target project's own linting commands and accessibility scanners to validate the quality of the UI code.
- [INDIRECT_PROMPT_INJECTION]: The skill processes untrusted content (source code, DOM text, and UI renders) from the project being reviewed, which represents an indirect injection surface.
- Ingestion points: Source code, DOM text, and UI screen captures are ingested from the project being reviewed (SKILL.md, Steps 1 and 3).
- Boundary markers: There are no explicit boundary markers or instructions to ignore embedded prompts within the reviewed content.
- Capability inventory: The skill has the capability to execute shell commands (linting, capture scripts) and write files to the local
.ui-review/directory (SKILL.md, Step 3 and Output section). - Sanitization: No explicit sanitization or validation of the ingested source code or DOM structure is performed before analysis.
Audit Metadata