What problem does it solve?
AI agents often struggle with complex, multi-step tasks, losing context, repeating work, or failing to track progress. This skill provides a structured framework to manage hierarchical work, ensuring clarity, traceability, and efficient task execution, allowing AI to work autonomously and reliably.
Core Features & Use Cases
- Hierarchical Work Cells: Organize tasks into versioned folders (
{branch}_v{version}_{step}) with DISCOVERY.md and STATE.md for clear context and progress tracking.
- State Synchronization: Explicitly mark work cells as
SYNCED or UNSYNCED to ensure accurate state reporting and prevent conflicting updates, making handoffs seamless.
- Versioned Restarts & Delegation: Create new versions (
v2) for fundamental rethinking or delegate sub-tasks to nested work cells, maintaining a clean history and enabling complex project management.
- Use Case: An AI agent is tasked with building a complex web application. It uses this skill to create separate work cells for
auth_v1_01, api_v1_01, and frontend_v1_01, each with its own STATE.md to track progress, and then delegates auth_v1_01/testing_v1_01 to a sub-agent, ensuring all work is organized and traceable.
Quick Start
Use yggdrasil-framework skill. Create a new work cell to achieve the following task: Use Python to write the first 10 fib numbers to a file.
