flutter-multi-environment
Pass
Audited by Gen Agent Trust Hub on Oct 24, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill instructions create a surface for indirect prompt injection as they require the AI agent to ingest and act upon data from untrusted external sources (the target repository's code and configuration files).
- Ingestion points: The agent is instructed to read project-specific files including
pubspec.yaml,pubspec.lock,android/app/build.gradle,ios/Runner.xcodeproj, andlib/main.dartto determine existing architecture and configuration. - Boundary markers: The instructions do not specify the use of delimiters or 'ignore embedded instructions' warnings when the agent processes these external files.
- Capability inventory: The agent has the capability to modify the filesystem (e.g., migrating entry points, creating new configuration files) and execute shell commands using the Flutter toolchain (
flutter run,flutter build,flutter test,flutter analyze). - Sanitization: The skill mitigates this risk by providing a robust validation framework in the
EnvironmentConfigclass, which includes strict parsing for environment names, URI schemes (enforcing HTTPS for production/staging), and boolean values to prevent malformed or malicious configuration from being used at runtime.
Audit Metadata