generate
Pass
Audited by Gen Agent Trust Hub on Sep 25, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided data via
$ARGUMENTS, including user stories, feature descriptions, and potentially external URLs. This data is used to generate test scripts that are subsequently executed in the local environment. - Ingestion points:
$ARGUMENTSfield (SKILL.md, Step 1). - Boundary markers: None explicitly defined to isolate untrusted user input within the generated code templates.
- Capability inventory: File writing (test files, page objects), file reading (source code, configs), and terminal execution via
ExecutionProxy/npx(SKILL.md, Step 7). - Sanitization: The instructions mention redacting secrets/PII before sharing context with subagents, but do not specify sanitization of the generated code itself to prevent injection.
- [COMMAND_EXECUTION]: The skill is designed to execute shell commands to verify tests.
- Evidence: Uses
rtk npm testandnpx playwright test <generated-file>to run the generated test suites (SKILL.md, Step 7). While this is intended functionality for a test generation tool, it creates a surface where generated (and potentially influenced) code is executed locally.
Audit Metadata