Opens AI-generated plans in a dedicated TUI for line-by-line user review and granular feedback collection.
The lgtm-plan skill enables a robust human-in-the-loop workflow by allowing developers to review AI-generated execution plans within a specialized Terminal User Interface (TUI). By facilitating a detailed line-by-line inspection, the skill ensures that developers can catch potential issues or provide specific guidance before Claude executes a proposed plan. Once the review session is closed, all feedback is automatically exported and read back into the conversation, allowing Claude to refine its strategy based on precise human input.
Key Features
01Interactive Terminal User Interface (TUI) for plan inspection
02Automatic synchronization of user feedback back to the Claude session
03Tmux split support for seamless side-by-side terminal workflows
04Support for both local development and marketplace installations
054 GitHub stars
06Line-by-line feedback and annotation capabilities
Use Cases
01Safety-checking automated batch file operations or scripts
02Providing granular corrections to architectural migration plans
03Verifying complex refactoring steps before AI implementation
