dart-language
Pass
Audited by Gen Agent Trust Hub on Sep 7, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [SAFE]: The skill provides coding guidelines and migration instructions for Dart 3 features, focusing on records, patterns, and sealed classes. These are standard development instructions.
- [SAFE]: All external links point to official Dart and Flutter documentation domains (dart.dev, flutter.dev), which are recognized as trusted organizations.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external project data, including source code and
pubspec.yamlconfiguration files. While this creates an ingestion surface for potentially untrusted content, the skill's capabilities are limited to benign operations such as static analysis, formatting, and documentation, posing no significant security risk. - Ingestion points: Reads project source code,
pubspec.yaml, and project CLI outputs within the local environment. - Boundary markers: None explicitly defined to delimit untrusted source code from instructions.
- Capability inventory: Uses Dart MCP semantic lookup, formatting tools, and static analysis; no arbitrary command execution or network exfiltration of analyzed data is described.
- Sanitization: Standard for static analysis tools; no specific sanitization of code content before processing is mentioned.
Audit Metadata