react-native-best-practices

Pass

Audited by Gen Agent Trust Hub on Nov 4, 2026

Risk Level: SAFEREMOTE_CODE_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
  • [REMOTE_CODE_EXECUTION]: The skill provides instructions to download and apply patches to core build-time dependencies (metro and metro-runtime). These patches are fetched from the official Software Mansion GitHub repository (https://raw.githubusercontent.com/software-mansion/react-native-reanimated/) and applied via git apply or patch. As these resources originate from the skill's vendor and target official framework repositories, they are documented as safe behavior for configuring advanced worklet features.
  • [DYNAMIC_EXECUTION]: The skill provides detailed guidance on using the React Native JSI (JavaScript Interface) and react-native-worklets, which include primitives for executing code dynamically. This includes jsi::Runtime::evaluateJavaScript and various worklet scheduling methods like runOnUISync and runOnRuntimeSync. These are documented as core features of the libraries for high-performance native-JS interaction.
  • [INDIRECT_PROMPT_INJECTION]: The skill provides tools for rendering external HTML and Markdown data. 1. Ingestion points: EnrichedText and EnrichedMarkdownText components in references/rich-text/SKILL.md. 2. Boundary markers: The documentation explicitly warns that the built-in useHtmlNormalizer is not a security sanitizer. 3. Capability inventory: The agent has access to instructions for JSI C++ bindings and Worklet runtimes which provide powerful native capabilities. 4. Sanitization: The skill instructs that user-provided HTML from external sources must be whitelisted and validated manually.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 4, 2026, 08:11 PM