flutter-health-audit
Fail
Audited by Gen Agent Trust Hub on Nov 16, 2026
Risk Level: CRITICALCOMMAND_EXECUTIONEXTERNAL_DOWNLOADSINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes various Bash commands for environment setup and project analysis. This includes installing tools like FVM, running Flutter tests with coverage flags, and using
findandgrepto analyze the repository structure and code patterns. These operations are restricted to the local environment and the target project's files. - [EXTERNAL_DOWNLOADS]: The
references/tool-installer.mdfile contains instructions to installNode.jsvia Homebrew andFVMvia the Dart package manager (pub). These are standard, well-known developer tools sourced from official registries, which is considered safe behavior for a development-focused skill. - [INDIRECT_PROMPT_INJECTION]: The skill exhibits an indirect prompt injection attack surface as it processes untrusted data from the repository being audited (e.g., README files, AI harness configurations, and source code).
- Ingestion points: The skill reads
pubspec.yaml,analysis_options.yaml,CLAUDE.md,.claude/rules/*.md, and other project files into the agent's context. - Boundary markers: The instructions emphasize evidence-based reporting but do not specify the use of strict delimiters or sanitization for content extracted from these files.
- Capability inventory: The skill possesses significant capabilities including
Bashfor command execution andAgentfor sub-task delegation, which could potentially be manipulated by malicious instructions hidden in the audited project files. - Sanitization: No explicit sanitization or filtering of external content before prompt interpolation is documented.
Recommendations
- CRITICAL: 2 file(s) identified as malware by FileRep - DO NOT USE
Audit Metadata