Skills
Navigate
skills/zakariaf/reactnative-skills/state-management-zustand/Gen Agent Trust Hub

state-management-zustand

Pass

Audited by Gen Agent Trust Hub on Oct 18, 2026

Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [COMMAND_EXECUTION]: The selftest.mjs script utilizes child_process.spawnSync to execute the skill's own analysis scripts (check-stores.mjs and check-reducers.mjs) against fixture directories. This is an intended internal testing mechanism and does not execute arbitrary user-supplied commands.
  • [DYNAMIC_EXECUTION]: The check-reducers.mjs script dynamically loads and executes user-defined reducer functions using Node.js's import() and loader hooks defined in app-modules.mjs. This capability is used to perform runtime correctness checks (e.g., detecting state mutations via deep freezing) and is constrained to the repository path provided by the user.
  • [INDIRECT_PROMPT_INJECTION]: The analysis scripts ingest source code files from the user's repository via walkRepo. This represents a potential ingestion surface for indirect instructions. However, the data is processed by local Node.js scripts for quality checks and is not passed to an LLM context in a manner that would trigger instruction following.
Audit Metadata
Risk Level
SAFE
Analyzed
Oct 18, 2026, 11:22 AM