flutter-platform-integration
Pass
Audited by Gen Agent Trust Hub on Oct 29, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is purely instructional and contains no scripts, binaries, or automated shell commands. It serves as a documentation resource for developers.
- [SAFE]: The skill provides explicit instructions to improve security, including treating external data (clipboard, deep links, URLs) as untrusted, enforcing data minimization, and keeping secrets (keystores, signing keys) out of the repository.
- [SAFE]: All external references point to official Flutter and Dart documentation on flutter.dev and pub.dev, which are well-known and trusted technology service domains.
- [INDIRECT_PROMPT_INJECTION]: The skill explicitly addresses the ingestion of untrusted data (clipboard, notifications, files, and URLs). It provides mitigation strategies, such as treating all incoming data as untrusted and implementing strict validation, ensuring the agent remains secure when handling external inputs.
- [DYNAMIC_EXECUTION]: The documentation recommends the use of Pigeon for typed contract generation. This is a standard build-time developer tool for Flutter and does not involve runtime dynamic code execution within the skill itself.
Audit Metadata