AI Unified Process - stack-agnostic core methodology (requirements, entity model, use cases)
93
95%
Does it follow best practices?
Impact
93%
1.43xAverage score across 10 eval scenarios
Passed
No known issues
Your team has finished the product discovery phase for TeamFlow, a new web-based task management platform. The product owner has written a vision document capturing the intended users, key features, and quality goals for the system. You can find this document at docs/vision.md.
Before development can begin, the engineering team needs a formal requirements catalog that structures the vision into actionable, verifiable requirements. The catalog will be used by developers to scope individual features, by QA to design test cases, and by the project manager to track delivery status.
A standards document is available in the project's references folder — consult it for the required ID formats, NFR categories, allowed data types, and quality checks to apply before finalizing the catalog.
Read docs/vision.md and produce a complete requirements catalog saved at docs/requirements.md.
The catalog should cover:
Each requirement type must be presented in its own dedicated Markdown table with appropriate columns. Every requirement row must be fully filled in — no empty cells.
Once drafted, review the catalog for quality before saving the final version: check that every requirement is clear, testable, and unambiguous, that all identifiers are unique across the document, and that every row's status is recorded.