dart-cognitive-complexity
Pass
Audited by Gen Agent Trust Hub on Aug 2, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill uses standard, read-only search commands (grep) to identify code patterns for analysis.
- [SAFE]: The verification step utilizes official Dart SDK tools (dart format, dart analyze) which are appropriate for the skill's stated purpose.
- [SAFE]: The instructions for subagent delegation explicitly include read-only mode guardrails to prevent unintended workspace mutations during audit sweeps.
- [SAFE]: No obfuscation, data exfiltration, or prompt injection patterns were detected in the instructions.
- [PROMPT_INJECTION]: The skill possesses an indirect prompt injection surface as it ingests untrusted Dart/Flutter code and provides instructions to execute it via 'dart test'. Ingestion points: User-provided lib/ and _test.dart files. Capability inventory: Execution of 'dart test' subprocess. Boundary markers: Absent for test execution context. Sanitization: None applied to ingested code before verification.
Audit Metadata