Start with the fields you know for sure: the core goal, the expected outcome, the audience or use case, and any constraints that affect the input. Optional fields can be left blank, but note what is unknown so the research backlog stays visible rather than hidden.
Tool entry input guide
Research-led guidance for shaping tool entry inputs clearly, handling missing fields, and keeping the next step easy to compare.
See overviewTool entry input guide
Format for clarity
Use short labels, one idea per field, and plain language that can be scanned quickly during review.
Avoid common mistakes
Do not mix goals with constraints, leave vague placeholders, or compress several assumptions into one input line.
Compare strong vs weak
Strong inputs are specific, bounded, and complete enough to test; weak inputs are broad, ambiguous, or missing the main context.
Common questions
What if some fields are still unknown?
Mark them explicitly as open and continue with the confirmed fields, so the input remains usable without pretending the research is complete.
How much detail is enough?
Include enough detail to reduce interpretation errors, but stop before adding unnecessary background that does not change the next action.
What should be researched next?
Validate the missing requirements, compare the inputs for consistency, and check whether any field needs a clearer definition before moving on.