write-plan
Pass
Audited by Gen Agent Trust Hub on Sep 18, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill ingests user-provided specifications to generate implementation plans that include code and shell commands. A malicious specification could attempt to inject instructions designed to be executed during the implementation phase.
- Ingestion points: The skill takes a 'spec or requirements' for a task as input in SKILL.md.
- Boundary markers: No explicit delimiters or 'ignore embedded instructions' warnings are defined for the input specification.
- Capability inventory: The generated plans contain shell commands like 'git commit' and 'pytest', which are executed via sub-skills like 'executing-plans' or 'subagent-driven-development' (SKILL.md).
- Sanitization: No explicit sanitization or validation of the input specification content is mentioned.
- [SAFE]: The skill incorporates defensive design by requiring a threat model review using the '/cso' command for any feature involving authentication, sensitive data, or trust boundaries.
- [SAFE]: The workflow enforces standard engineering best practices including TDD, SOLID, DRY, and YAGNI, and requires a complete verification and test plan artifact.
Audit Metadata