recipe-fullstack-implement
Pass
Audited by Gen Agent Trust Hub on Mar 13, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [PROMPT_INJECTION]: The skill exhibits an indirect prompt injection vulnerability surface due to its orchestration of external data. \n- Ingestion points: The skill ingests untrusted data via the
$ARGUMENTSvariable and user-provided prototype code paths. \n- Boundary markers: It attempts to mitigate injection by mandating a[SYSTEM CONSTRAINT]suffix for all sub-agent prompts, though it lacks formal delimiters for user-supplied data. \n- Capability inventory: The orchestrator invokes sub-agents (e.g.,task-executor,quality-fixer) that possess capabilities for code modification and task execution. \n- Sanitization: There is no evidence of validation, escaping, or schema enforcement for user-provided strings before they are interpolated into sub-agent prompts.
Audit Metadata