test-cases
Pass
Audited by Gen Agent Trust Hub on Oct 9, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill is primarily focused on generating test case documentation (UT/IT/E2E/Journey) based on project requirements. It operates within a strictly defined documentation phase and is explicitly prohibited from producing implementation or test code (e.g., '⛔ 禁止继续：文档阶段不得产出实现代码或测试代码文件'), which significantly limits the potential for malicious behavior.
- [INDIRECT_PROMPT_INJECTION]: The skill ingests untrusted data from 'docs/devdocs/01-requirements.md' to drive its documentation output. While this presents an indirect prompt injection surface, the risk is categorized as safe given the lack of executable capabilities and the presence of strict phase-boundary constraints.
- Ingestion points: 'docs/devdocs/01-requirements.md' (via Read tool).
- Boundary markers: No explicit 'ignore embedded instructions' delimiters are defined for the input requirement data.
- Capability inventory: Access to 'Read', 'Write', 'Glob', 'Grep', and 'AskUserQuestion'. No network or shell execution tools are enabled in frontmatter.
- Sanitization: No explicit sanitization or escaping of external requirement content is performed before interpolation into markdown templates.
Audit Metadata