NYC

react-native-best-practices

Pass

Audited by Gen Agent Trust Hub on Feb 16, 2026

Risk Level: LOWPROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • INDIRECT_PROMPT_INJECTION (LOW): The skill is designed to analyze untrusted external content in the form of user-provided React Native code.\n
  • Ingestion points: User-provided React Native source code processed according to guidelines in SKILL.md.\n
  • Boundary markers: Absent. There are no delimiters or instructions to ignore instructions embedded in the analyzed code.\n
  • Capability inventory: The skill references shell commands (grep, npx in SKILL.md) for code analysis and search, which could be targeted if an agent executes these on malicious source code.\n
  • Sanitization: Absent. No filtering or validation of external code is specified.\n- COMMAND_EXECUTION (LOW): The skill contains informational bash and npx command examples (e.g., npx source-map-explorer) for performance profiling. These are standard development tools and do not involve suspicious obfuscation or malicious intent.\n- EXTERNAL_DOWNLOADS (INFO): Mentions several reputable React Native ecosystem packages (e.g., jotai, zustand, react-native-performance) as optimization recommendations. No untrusted remote scripts or download-to-pipe patterns were found.
Audit Metadata
Risk Level
LOW
Analyzed
Feb 16, 2026, 09:41 AM