gsd-research-phase

Implementation research for project management investigates standard tech stacks, architecture patterns, and common pitfalls before committing to a full executable plan. This process validates phase feasibility and gathers necessary context to reduce downstream planning rework.

The best way to research implementation approaches is to spawn a dedicated isolated researcher agent that investigates standard libraries and common gotchas. This avoids consuming your main conversation context while producing structured research documentation.

Yes, you can update existing phase research. The skill detects existing research files and lets you choose to update, view, or skip them to avoid redundant work and prevent duplicate research efforts.

Yes, phase validation automatically checks that the target phase exists in your project roadmap before starting research. This ensures research effort is only spent on valid, planned phases.

You gather implementation context without burning main context by spawning an isolated researcher agent with fresh context. This dedicated agent investigates patterns and pitfalls independently and outputs structured documentation.

A key limitation is that the research output must align with downstream GSD planning tooling requirements. Additionally, the skill assumes your target phase already exists in a defined project roadmap before research can begin.