Please show us one ordinary input
#7
by ArjunNaidu - opened
The natural-language input sounds approachable. An example of a plain, slightly messy request would help more than a perfectly polished prompt. It would be useful to see which details MatterSeek extracts and which ones it asks users to clarify.
We've actually documented the full parsing flow in our help docs: a plain, unstructured request → the properties, ranges and constraints MatterSeek extracts automatically → the specific points where it pauses and asks the user to clarify (e.g., ambiguous property names or missing ranges).
You can find it in the help documentation — and if anything there feels unclear after you try it, let us know. That "what we understood vs. what we asked back" transparency is exactly the standard we want to keep.