qe-test-execution
Pass
Audited by Gen Agent Trust Hub on Mar 14, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits a vulnerability surface for indirect prompt injection.\n
- Ingestion points: Data enters the agent context via Pull Request analysis (e.g.,
PR #123) and the inspection of test files within the repository (SKILL.md).\n - Boundary markers: The provided task templates do not implement explicit boundary markers or instructions to disregard embedded commands in the analyzed data.\n
- Capability inventory: The skill possesses the capability to execute system commands via the
aqeCLI tool (SKILL.md).\n - Sanitization: There is no evidence of sanitization, escaping, or validation of the untrusted PR descriptions or file content before processing.\n- [COMMAND_EXECUTION]: The skill orchestrates the execution of the
aqeCLI tool for running test suites. This involves constructing and executing shell commands with various flags such as--parallel,--workers, and--shard.\n- [EXTERNAL_DOWNLOADS]: The CI/CD integration example utilizes well-known GitHub Actions (actions/checkoutandactions/upload-artifact), which are considered trusted sources.
Audit Metadata