subagent-driven-development

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill consumes implementation plans from external files to drive subagent actions. 1. Ingestion points: The process involves reading plan files to extract task descriptions. 2. Boundary markers: Subagent templates in 'implementer-prompt.md' and 'spec-reviewer-prompt.md' use Markdown headers and specific placeholders to delimit task text from instructions. 3. Capability inventory: Implementer subagents are granted permissions to modify files, run tests, and commit code. 4. Sanitization: The workflow relies on independent multi-stage review subagents (spec compliance and code quality) to validate output and detect issues introduced during implementation.
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 09:01 AM
Security Audit — agent-trust-hub — subagent-driven-development