tdd-workflows-tdd-cycle
Pass
Audited by Gen Agent Trust Hub on Oct 17, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface due to the way it handles user-provided data.
- Ingestion points: User-supplied input via
$ARGUMENTSis interpolated into multiple prompts for subagents such asarchitect-review,test-automator,code-reviewer, andbackend-architectinSKILL.md. - Boundary markers: The skill lacks explicit boundary markers (e.g., XML tags or distinct delimiters) or negative constraints to prevent subagents from obeying instructions embedded within the
$ARGUMENTScontent. - Capability inventory: The orchestrated subagents are granted significant capabilities, including designing architectures, writing executable tests, implementing production code, and performing code refactoring, which involves extensive file system interaction and code modification.
- Sanitization: There is no evidence of input validation, escaping, or sanitization performed on the
$ARGUMENTSvariable before it is passed to the task orchestration tool.
Audit Metadata