Plan Mode

What is Plan Mode?

Plan mode is a human-AI deep collaboration development mode designed for complex tasks. It mimics the natural behavior of human expert teams when solving complex requirements, emphasizing early consensus building, conducting in-depth discussions, analysis, and research based on requirements, formulating systematic technical design plans, and then executing through division of labor.

In plan mode, a "Research-Action-Review" three-stage process model is introduced, encouraging more time to be invested in collaborative planning in the early stages, significantly reducing rework costs caused by understanding deviations later. AI transitions from a tool to a partner, with users responsible for domain knowledge and key decisions, while AI handles efficient execution and robust generation, achieving efficient and controllable R&D intelligence.

You can describe tasks in natural language, and AI will first clarify requirements and generate an initial research plan based on a deep understanding of the project. Then, using deep search and tool invocation capabilities, it conducts comprehensive research and outputs standardized user stories (requirement.md) and technical designs (design.md), greatly improving document quality and credibility. Subsequently, it autonomously plans and breaks down the documentation into executable and trackable detailed development tasks (To-dos) and executes them efficiently.

Concept Terminology

Concept	Definition
Research Plan	A coarse-grained research outline generated by clarifying goals and boundaries through questioning based on the user's original input.
User Stories (requirement.md)	Structured user stories that transform user input into clear, complete, testable/acceptable, and traceable requirements.
Technical Design (design.md)	An overall overview of complex project engineering, including system architecture, key implementations/components/modules, and sequence diagram interactions.
Development Tasks (To-dos)	Independently executable and trackable development tasks, each with a clear definition of description, expected results, and necessary dependencies.

Research Artifacts

Research artifacts ensure consistency and reduce development iterations:

• User Stories (requirement.md): Structured user stories that transform user input into clear, complete, testable/acceptable, and traceable requirements.

• Technical Design (design.md): An overall overview of complex project engineering, including system architecture, key implementations/components/modules, and sequence diagram interactions.

• Development Tasks (To-dos): Independently executable and trackable development tasks, each with a clear definition of description, expected results, and necessary dependencies.

Using Plan Mode

Open the AI conversation sidebar homepage to switch to plan mode.