Document Task Planner

Overview

This agent acts as a specialized task planner designed to ingest unstructured or semi-structured documents (like feature specs, bug reports, or design docs) and transform them into a highly actionable, structured task manifest. It is crucial for kickstarting development sprints by ensuring every required piece of work is isolated into discrete, executable tasks.

Capabilities

• Actionable Item Extraction: Reads source documents to identify all concrete items that require implementation or completion.
• Structured Formatting: Outputs tasks in a strict markdown format (.dispatch/tasks-{name}.md) adhering to specific dependency and prompt structures.
• Task Granularity Enforcement: Breaks down large concerns into the smallest possible, single-focus tasks (e.g., separating parsing from interpretation).
• Dependency Mapping: Accurately maps task dependencies (T1 must finish before T2 starts) while maximizing parallelism where possible.

Example Use Cases

Imagine receiving a 50-page feature specification document. Instead of reading it linearly, you feed it to this agent. It will return a file containing dozens of tasks: one for setting up the database schema, several for implementing specific API endpoints (each isolated), and separate tasks for writing unit tests and updating documentation. This ensures that development work can be tracked, assigned, and completed module by module, preventing scope creep or missed steps.