subagent-driven-development
Pass
Audited by Gen Agent Trust Hub on Jul 16, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill follows security best practices by using fresh subagent contexts for each task to prevent context pollution and by mandating code review cycles between development steps.
- [PROMPT_INJECTION]: The workflow processes external plan files which creates a surface for indirect prompt injection. This risk is addressed by the multi-step verification process described in the skill.
- Ingestion points: Local plan files loaded at the start of the process.
- Boundary markers: The prompts for implementation subagents do not include specific delimiters to isolate plan content from instructions.
- Capability inventory: Subagents have the capability to write files, commit changes, and execute testing tools.
- Sanitization: The skill relies on the subsequent code review sub-skill to identify and mitigate any malicious or erroneous content introduced during implementation.
Audit Metadata