subagent-driven-development

Pass

Audited by Gen Agent Trust Hub on Mar 14, 2026

Risk Level: SAFEPROMPT_INJECTIONNO_CODE
Full Analysis
  • [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection. Ingestion points: Task text from implementation plans is passed into subagent prompts. Boundary markers: Markdown headers are used for separation. Capability inventory: Subagents have the ability to write code and execute tests. Sanitization: No sanitization or escaping is applied to the injected task text.
  • [NO_CODE]: This skill consists solely of markdown prompt templates and documentation with no embedded executable scripts.
Audit Metadata
Risk Level
SAFE
Analyzed
Mar 14, 2026, 11:05 AM