Hi i'm trying to get AI to analyse images sent in by users via WhatsApp. I've got access to a new card under 'Agents' called "Extract Content from Image" which seems like it would do the job. However... it seems to require me to insert a url of the image to analyse, which in my case will be unknown (because the user sends the image in) Anyone have any experience with this kind of use case? Ps Warning in advance. I've recently discovered that I am basically code illiterate. https://cdn.discordapp.com/attachments/1231982274829094945/1231982275046932634/Screenshot_2024-04-22_at_15.49.06.png?ex=6638efe7&is=66267ae7&hm=61fa4538493ceac53438cf5efb48d4461ce36f695ab3dfd1c0f8cb51e14561a3& https://cdn.discordapp.com/attachments/1231982274829094945/1231982275273560145/Screenshot_2024-04-18_at_15.10.48.png?ex=6638efe7&is=66267ae7&hm=4970ed5b005a9770e59ba2f47fc18da467345f86519845a40d496ef7f20b8433&

Hey there, You can include a variable in the image url part. I haven't tried it, but I am sure that someone like @bumpy-addition-21507 or @limited-pencil-78283 has done something similar before. https://botpress.com/docs/cloud/channels/whatsapp/#media-images-audio-documents

Still haven't given up 💪 I went back to basics and am currently now struggling to get a response when using a url. Do these logs give any insight as to why? https://cdn.discordapp.com/attachments/1231982274829094945/1232285157969625148/Screenshot_2024-04-23_at_11.58.59.png?ex=6628e67c&is=662794fc&hm=9b2561b58949bfb2ab296b575872892b0ea92affd4623d4a5fb5815831f7b806& https://cdn.discordapp.com/attachments/1231982274829094945/1232285158372282430/Screenshot_2024-04-23_at_11.58.39.png?ex=6628e67c&is=662794fc&hm=20d81b989ce0f0cb867179e0dd171119066adbf19604dd86db129ee02ef7dede& https://cdn.discordapp.com/attachments/1231982274829094945/1232285158711890032/Screenshot_2024-04-23_at_11.58.20.png?ex=6628e67c&is=662794fc&hm=716c15c6a0b4fa6b6248a85be2a6fb84571d271fd26d888fcf8f02f3d91d4aaa&

Probs a bad url

What would make a url good or bad? I'm now trying to have the image analysis done via a Zap in Zapier. The rest done via Botpress still

It has to be publicly available

Ok, got it. Thanks. I've actually now made it so that the url input field on the card "extract content from image" to work when there is a variable put in it, instead of just a url. Now I just need to save/store the incoming WhatsApp image as a variable somehow?

Remember to use the scope like workflow or user So it would be @workflow.testimage if the scope is workflow

Well spotted, thanks. I'm getting there slowly. Hopefully i'll be the worlds expert on whatsapp image saving soon 😂

I am sure that you will!🛠️

@fresh-fireman-491 - I have been working in the above approach and seems like little problem. I have created a variable with Userinputurl for extract content from image card and passing this value to the card and storing the output with other output variable "image". But when i use the output variable in the text card it says value as "undefined". but strangely when i use the input variable "Userinputurl" it is showing the answer which is expected.

That sounds weird. Could you create a new post in #1111009377525186570 and share some more information including examples