Work Pipeline Scheduler

Overview

This agent acts as the central orchestrator for complex, multi-stage workflows (WORKs). It reads a defined Task Dependency Graph (DAG) associated with a specific WORK and ensures that all prerequisite tasks are completed before proceeding. It manages the entire execution sequence, dispatching results sequentially through builder, verifier, and committer subagents.

Capabilities

• Dependency Resolution: Analyzes the TASK dependency DAG to determine which tasks are ready for execution based on completion status.
• Sequential Dispatch: Calls specialized agents (builder $\rightarrow$ verifier $\rightarrow$ committer) in the correct, dependent order for each task.
• State Tracking: Maintains and updates progress records (PROGRESS.md) throughout the entire pipeline run.
• Error Handling: Implements retry logic (up to 3 times) when a builder step fails.
• Logging & Reporting: Logs every stage of execution into a dedicated work log file and provides status updates upon completion.

Example Use Cases

1. Full Feature Implementation: When you have defined a complex feature that requires initial code generation (builder), subsequent quality checks (verifier), and final integration commits (committer).
2. Automated Testing Cycles: Running an entire suite of dependent tests where one test must pass before the next can begin.
3. Multi-Step Data Processing: Executing a data pipeline where cleaning, transformation, and loading must occur in strict sequence.