expo-brownfield

Pass

Audited by Gen Agent Trust Hub on Aug 26, 2026

Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [COMMAND_EXECUTION]: The skill provides build configuration templates (Gradle scripts and iOS Podfiles) that execute shell commands at build time to resolve package paths and execute build phases.
  • Evidence: settings.gradle and app/build.gradle in references/brownfield-integrated.md use providers.exec and commandLine to run node for resolving paths to @react-native/gradle-plugin and expo-modules-autolinking.
  • Evidence: The iOS Podfile in references/brownfield-integrated.md uses shell execution (backticks) to resolve the path for expo/package.json.
  • Evidence: A custom Xcode Run Script phase in references/brownfield-integrated.md executes shell logic to embed the JavaScript bundle during the build process.
  • [EXTERNAL_DOWNLOADS]: The skill instructs users to utilize standard package managers to download and install official development tools and libraries.
  • Evidence: Instructions in SKILL.md and reference files recommend using npx create-expo-app, npx expo install, and yarn install to manage project dependencies.
  • Evidence: Documentation in references/brownfield-integrated.md mentions sudo gem install cocoapods to install the iOS dependency manager.
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process user context regarding their specific native application architecture to provide tailored integration advice.
  • Ingestion points: The skill triggers based on user keywords such as "brownfield", "AAR", and "XCFramework" (SKILL.md).
  • Boundary markers: Absent.
  • Capability inventory: The skill produces shell commands for project initialization, package installation, and build configuration (references/brownfield-integrated.md, references/brownfield-isolated.md).
  • Sanitization: Absent.
Audit Metadata
Risk Level
SAFE
Analyzed
Aug 26, 2026, 06:06 AM