flow-next-impl-review
Pass
Audited by Gen Agent Trust Hub on May 13, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill's architecture and behavior are consistent with its purpose as a professional development tool. It uses bundled scripts and communicates with well-known AI services without exhibiting any malicious patterns or unauthorized behaviors.
- [COMMAND_EXECUTION]: The skill relies on a bundled utility named
flowctland standard system commands (git, jq) to analyze the repository, manage review receipts, and coordinate fixes. This is appropriate for a repository-integrated review automation tool. - [EXTERNAL_DOWNLOADS]: The skill performs network operations to interact with well-known AI backends including GitHub Copilot, OpenAI, and RepoPrompt. These connections are intrinsic to the backends described in the documentation and are used solely for code analysis.
Audit Metadata