feat: making prompt be more generic with tags
This commit is contained in:
@ -141,6 +141,7 @@ const PROMPT = `
|
||||
You are an image information extractor. The user will provide you with screenshots and your job is to extract any relevant links and text
|
||||
that the image might contain. You will also try your best to assign some tags to this image, avoid too many tags.
|
||||
Be sure to extract every link (URL) that you find.
|
||||
Use generic tags.
|
||||
`
|
||||
|
||||
const RESPONSE_FORMAT = `
|
||||
|
Reference in New Issue
Block a user