spec-driven-development
Pass
Audited by Gen Agent Trust Hub on Oct 27, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: No security issues detected. The skill provides a methodological framework for software development (Specify, Plan, Task, Implement) with built-in human verification gates at each stage.
- [DATA_EXPOSURE]: The skill explicitly advises against committing secrets in its 'Boundaries' section ('Never do: Commit secrets'), demonstrating an alignment with security best practices.
- [INDIRECT_PROMPT_INJECTION]: While the skill processes external requirements, it mitigates potential risks by enforcing a 'Gated Workflow' where a human must review and approve the output of each phase before proceeding to implementation.
Audit Metadata