subagent-driven-development
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The workflow involves ingesting external implementation plans which are then processed by subagents.
- Ingestion points: Task descriptions from plan files are interpolated into subagent prompts in implementer-prompt.md and spec-reviewer-prompt.md.
- Boundary markers: The skill uses markdown sections to delineate requirements, though no specific escaping of task content is mentioned.
- Capability inventory: Subagents are expected to write code, perform file operations, and execute tests as part of the implementation tasks.
- Sanitization: The skill mitigates risks through a two-stage review process (spec compliance and code quality) performed by separate subagent instances.
Audit Metadata