frontend-state-and-form-management
Pass
Audited by Gen Agent Trust Hub on Sep 21, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill instructions provide legitimate architectural patterns for frontend development, specifically focusing on React state, Zustand stores, and TanStack Query. The directives to prioritize source code generation over text descriptions are standard for developer-oriented tools.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest external requirements ('Analysis Plan', 'API contract') to drive code generation, which represents a standard vulnerability surface for coding assistants.
- Ingestion points: External requirements for business validation rules and persona-specific logic (SKILL.md, sections 5 and 6).
- Boundary markers: The instructions do not specify delimiters or explicit warnings for handling embedded instructions within these external requirements.
- Capability inventory: The skill directs the agent to perform file system write operations for source code generation (SKILL.md, 'WRITE FILES, NOT DESCRIPTIONS').
- Sanitization: Input data is processed as logical requirements for the agent to translate into code, rather than being directly executed.
Audit Metadata