validate-ui-scenario
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- Command Execution: The skill executes shell commands such as
npm installandnodeto compile test scenarios and run the UI automation harness. - Dynamic Code Execution: The skill generates JavaScript files (
.cjs) and executes them using a local Node.js runner to drive a VS Code instance. It also uses Playwright'spage.evaluate()to execute arbitrary JavaScript within the browser context for tasks like metadata collection from the UI. - Indirect Prompt Injection Surface: The skill is designed to ingest and act upon data from external sources, specifically GitHub issue descriptions, to generate the automation logic. This represents a surface where untrusted content could influence the generated script code.
- Ingestion points: Reads content from provided GitHub issue URLs to identify reproduction steps.
- Boundary markers: None explicitly defined to separate untrusted issue content from the code generation instructions.
- Capability inventory: Can write files to the local system, execute shell commands via Node.js, and interact with the file system through Playwright.
- Sanitization: No explicit sanitization or validation of the external issue text is performed before it is used to formulate the automation scenario.
Audit Metadata