test-coverage
Pass
Audited by Gen Agent Trust Hub on Jun 29, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to execute shell commands to run tests (e.g., unit, contract, integration, and E2E tests) and capture their output for reporting. This is the primary intended functionality of the skill for verifying product behavior.
- [PROMPT_INJECTION]: The skill processes untrusted external data, including PRDs, feature breakdowns, and source code changes, to generate test specifications. This creates an indirect prompt injection surface where malicious instructions in project documentation could influence the generated test cases. However, the risk is mitigated by the skill's reliance on structured Markdown templates and its focus on recording factual results rather than qualitative assessments.
- Ingestion points: SKILL.md (Workflow section) identifies reading PRDs, feature breakdowns, specs, changed implementation, and existing tests.
- Boundary markers: Uses Markdown-based templates (assets/test-spec-template.md, assets/test-report-template.md) to structure output.
- Capability inventory: Executes arbitrary shell commands for verification and drives external test producer skills.
- Sanitization: No explicit sanitization logic is described for the ingested text.
- [CREDENTIALS_UNSAFE]: The skill explicitly instructs the agent to never include passwords, API keys, cookies, tokens, or database URLs in any generated specs, reports, or logs, demonstrating a strong security posture regarding credential management.
Audit Metadata