aep-validation
Pass
Audited by Gen Agent Trust Hub on Aug 6, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSDATA_EXFILTRATIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill automates validation workflows by executing system commands including git for branch management, gh for issue and PR operations, npm for dependency installation, and npx for running Playwright tests. It also executes a local Node.js script, generate-report.mjs, which uses execFileSync to perform git operations.
- [EXTERNAL_DOWNLOADS]: Fetches the @playwright/test package using npm. It also performs network requests via curl to a platform-defined API ($AEP_PLATFORM_URL) to retrieve dynamic test credentials required for validation tasks.
- [DATA_EXFILTRATION]: Accesses a platform-provided bearer token file ($AEP_BEARER_FILE) to authenticate legitimate API calls to the platform's credential service. This is a standard part of the skill's integration with the host platform.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface. It ingests data from GitHub issue comments, the validation-criteria.json file, and validation-context.json. While it implements boundary markers by designating the criteria as read-only and performs basic sanitization in its report script, its extensive capabilities (including shell execution via git, gh, and curl) remain exposed to instructions potentially embedded in the processed data.
Audit Metadata