Recognises a guess recorded as fact
A gap in the source filled with a plausible value and no open_question. Also: a requirement reverse-engineered from an existing schema or sample data, recorded as a business need.
Can AI do this?
AI does this well Most of the tested AI models master this. 8 of 10 models master it.
| Model | Tier | Probability of mastery | Status |
|---|---|---|---|
| gemini-4-argon | Flagship | 100% | Mastered |
| gemini-3.1-pro-preview | Flagship | 0% | Not shown |
| gpt-6-astra | Flagship | 100% | Mastered |
| claude-fable-5-1 | Flagship | 100% | Mastered |
| gemini-3.8-flash | Mid tier | 94% | Mastered |
| gpt-6.1-sol | Mid tier | 100% | Mastered |
| claude-sonnet-5-5 | Mid tier | 100% | Mastered |
| gemini-3.5-flash-lite | Low cost | 0% | Not shown |
| gpt-6-luna | Low cost | 100% | Mastered |
| claude-haiku-5-5 | Low cost | 100% | Mastered |
Mastered means a probability of at least 90% that the model has this knobit, estimated from its graded answers. “Not shown” means the answers do not settle it either way. How the models are tested.
What it looks like in practice
Expected result: The guessed content and the open question it should be
How it is assessed: Flags the deficient case and passes the acceptable one. Deficient: "Payment references are unique." taken from a unique index in the legacy database, stakeholder: Payments Ops, status: confirmed Acceptable: source.document: "existing system: PAYMENTS_DB DDL", status: draft, rationale: "describes current behaviour, not confirmed as a business need", plus an open_question for Payments Ops.
Also known as: Guess recorded as fact (data requirement)
Related knobits
Builds on
All 47 knobits of Capture data requirements.
Source: Review checklist 01-requirements, row 19; Stage 1 (record what you did not get)