zustand-use

Pass

Audited by Gen Agent Trust Hub on Sep 15, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDYNAMIC_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill triggers file system operations based on user-supplied strings (e.g., "count 스토어 생성") to automate store creation.
  • Ingestion points: User-provided store names are extracted from natural language prompts as described in the "Store File Auto-generation" section.
  • Boundary markers: The skill does not explicitly define delimiters to separate user-provided names from potential instructions, though it directs the agent to interpret the user's intent.
  • Capability inventory: The skill utilizes the agent's ability to create directories (src/store) and write files to the local file system.
  • Sanitization: The skill mandates "Name normalization" (camelCase conversion) for the provided names, which acts as a basic filter for the generated identifiers and filenames.
  • [DYNAMIC_EXECUTION]: The skill generates boilerplate code for Zustand stores in both TypeScript and JavaScript using user-supplied names as variable and file identifiers.
  • The generation follows fixed templates provided in the skill instructions.
  • The logic includes environmental checks (looking for tsconfig.json) to determine the appropriate language (TS vs JS).
Audit Metadata
Risk Level
SAFE
Analyzed
Sep 15, 2026, 08:04 AM
Security Audit — agent-trust-hub — zustand-use