subagent-driven-development
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill templates interpolate external task data into subagent instructions, creating a surface for indirect prompt injection.
- Ingestion points: Task descriptions and requirements are pasted into
implementer-prompt.mdandspec-reviewer-prompt.md. - Boundary markers: The templates use standard Markdown headers but lack strong delimiters or explicit warnings to ignore embedded instructions within the task data.
- Capability inventory: The implementer subagent is granted capabilities to write code and execute tests (command execution) as part of its primary function.
- Sanitization: No explicit sanitization or validation of the input implementation plan text is implemented within the templates.
Audit Metadata