Planning Stages
Full projects follow a structured, stage-based delivery lifecycle. At each stage, an AI assistant acts as your collaborative thought partner — adapting its guidance to your role and the current phase of work.
The Journey Through Your SDLC
1. Explore & Write PRD
Who: Product Manager
This is where every project begins. You chat with the AI to explore the problem space — identifying target users, defining success criteria, and crystallizing your thinking into a formal PRD document.
The AI focuses on deep listening and exploration, helping you articulate the "what" and "why" before jumping to solutions. Implementation details are intentionally avoided at this stage.
2. Product Spec
Who: Product Manager
Once your PRD is ready, the AI analyzes it and proposes a breakdown into logical Workstreams — independent areas of work like "User Authentication" or "Payment Integration." You review and confirm the workstreams.
For each workstream, the AI generates detailed Requirements with acceptance criteria. Requirements start as Drafts and move to Confirmed as you finalize them.
Learn more about Workstreams & Requirements.
3. UI/UX
Who: Product Manager or UI/UX Designer
This stage translates requirements into tangible user experiences:
- Define Areas (logical sections of the UI) and Screens within them
- Map out User Flows for complex interactions
- Generate interactive HTML mockups directly in chat
- The AI grounds designs in your existing product patterns for consistency
Requirements without a UI component can be marked as "not needed" for this stage.
4. Engineering
Who: Engineering Manager, Tech Lead, or Developer
The technical blueprinting phase. The AI researches your actual codebase to ensure plans are realistic and grounded in existing patterns:
- Define technical architecture and API contracts
- Identify data model changes and dependencies
- Create implementation plans that build on your established conventions
5. Development
Who: Developers
This is where plans become code. Developers connect ProdE to their coding tools via MCP and work through requirements one by one:
- Requirements are broken into actionable development tasks
- Developers update requirement status as they implement (
Not Started→Planned→Developed) - The coding agent has full context — PRD, requirements, acceptance criteria, and engineering plans
See Connecting to Your Coding Agents for setup instructions.
6. QA
Who: QA Team
The final validation stage:
- Plan testing strategies based on acceptance criteria
- Track which requirements have been tested and verified
- Requirements move to
Testedstatus once validated
AI Adapts to Your Role
The AI tailors its guidance based on who you are:
| Role | The AI Focuses On |
|---|---|
| Product Manager | Business logic, user flows, product impact |
| Engineering Manager | Architecture, technical oversight, team coordination |
| Software Developer | Data models, APIs, implementation patterns |
| QA | Testing strategies, edge cases, quality assurance |
| UI/UX | User experience, wireframes, design patterns |
| Solution Consultant | Product capabilities, integration patterns |
| Support / Customer Success | Product behavior, troubleshooting |
The AI won't jump ahead or give irrelevant guidance — in the PRD stage it focuses on exploration; in Engineering it focuses on architecture grounded in your codebase.
Handling Mid-Cycle Changes
If requirements, designs, or technical plans need to change after a stage is complete, ProdE provides a structured workflow to:
- Explore the proposed change
- Assess its impact on existing artifacts
- Update affected documents and requirements systematically
When Is a Stage Complete?
| Stage | Complete When |
|---|---|
| PRD | Problem, users, solution approach, and success criteria are documented |
| Product Spec | All requirements are confirmed or marked as out of scope |
| UI/UX | All requirements have design status "completed" or "not needed" |
| Engineering | All requirements have dev status "planned" or later |
| Development | All requirements have been developed |
| QA | All requirements have been tested |
Next: Learn about Workstreams & Requirements for organizing and tracking work.