using-pm-team

To create a structured pre-development plan, use a gate-driven workflow that produces PRD, TRD, API design, data model, and dependency map artifacts before coding begins. This prevents ad-hoc planning and mid-implementation surprises by enforcing research-first requirements.

The best way to break down large software features is using a 9-gate Large Track planning workflow that produces dependency maps, feature maps, and zero-context subtasks. This hands off clear, structured engineering tasks sized for features requiring two or more days.

Yes, you need separate planning tracks for small and large features. A Small Track handles modifications under two days with 4 gates, while a Large Track manages features over two days with 9 gates, matching the planning depth to the feature's scope and risk.

Automated research fits into product planning by deploying research agents to analyze codebase patterns, web best practices, and framework documentation. This integrated research directly informs the PRD, TRD, and dependency mapping before technical design gates are completed.

Yes, you can generate API and data model designs as part of a pre-dev planning workflow. The process produces structured api-design and data-model documents alongside the TRD, ensuring technical specifications are fully mapped before implementation begins.

A limitation of gate-based planning tracks is the required completion of sequential artifacts like research.md, PRDs, and TRDs before coding. This strict enforcement prevents ad-hoc changes but may slow down immediate hotfixes requiring rapid, undocumented implementation.