run-smoke-tests
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to process untrusted test execution artifacts and apply automated code fixes, introducing an indirect prompt injection surface.
- Ingestion points: Test traces and application logs are inspected dynamically during step 3 of the workflow in
SKILL.md. - Boundary markers: Absent; there are no delimiters or explicit instructions telling the model to disregard malicious or manipulative text embedded inside error logs or test data.
- Capability inventory: File write/modification capabilities are implied by "Apply a minimal fix", and arbitrary command execution is available via
npm runcapabilities. - Sanitization: Absent; the skill does not filter or sanitize the log files or trace content before context interpolation.
Audit Metadata