What problem does it solve?
Executing complex multi-agent workflows can be opaque and difficult to manage, especially when errors occur or human intervention is needed. This Skill provides real-time visualization, interactive steering at checkpoints, and robust error recovery, making workflow execution transparent and fully controllable.
Core Features & Use Cases
- Real-time Visualization: See your workflow graph and progress updates as agents execute, providing immediate clarity on complex processes.
- Interactive Steering: Pause at checkpoints (
@review) to make critical decisions, retry steps, modify instructions, or abort, giving you full control over the workflow's direction.
- Graceful Error Recovery: If an agent fails, receive clear options to retry with the same or modified instructions, skip the step, or abort, preventing workflow halts.
- Automated Cleanup: Ensures all temporary files (scripts, agents, state) are deleted after every execution, maintaining a clean and efficient plugin workspace.
- Use Case: Execute a deployment pipeline that includes security scans and manual approval. If the security scan fails, the Skill will pause, offer options to fix and retry, or proceed with a fallback plan.
Quick Start
Run this workflow: Explore:'Analyze codebase' -> implement:'Add feature' -> general-purpose:'Run tests'
