I have an OpenAPI spec with the following parameter:
- in: query
name: limit
schema:
type: number
default: 3
description: Number of results to return. Always default to 3 unless specified
with an exact amount by the user.
required: false
I then use a prompt of the form “Show me some [resource I’m requesting]”. ChatGPT then consistently uses 5 as the limit
parameter instead of 3.
Is there anything else I could do to the spec to help encourage ChatGPT to help encourage the model to start with the defaults we set? Happy to create a minimal repo if my issue as stated is unclear.