Hi, I created a bot for a doctor who is both a general and aesthetic surgeon. I'm using an autonomous node and have provided all the necessary information in the knowledge base. When it comes to general information about treatments, everything works perfectly. However, I’m having issues with pricing. I uploaded a file containing the exact prices of all treatments and asked the bot to strictly follow that file. Despite this, the bot doesn't stick to the listed prices. Sometimes it gives the correct ones, but other times it provides prices that are not in the file at all. I’ve tried changing the file type (DOCX, PDF, Excel, etc.), but it doesn’t make a difference. How can I make sure the bot gives only the prices that are actually in the file? The problem is that I'm using an autonomous node, and I don’t know any other way to input this information than by uploading it to the knowledge base. Any suggestions? Thanks

Hi, a client wants that the Botpress Bot pops up, when a person is visiting the website. How could I do that? I guess there has to come a Trigger from the website, right?

I'm almost completely new to botpress, as well as javascript. My goal is to have the client upload a PDF, and the bot puts it in a certain folder in our knowledgebase that it can then reference information from. I was able to get the autonomous node to go into a secondary workflow where it asked for the file (file card), then executes this code. Unfortunately it gives says there was an error executing the action. checking the logs it says the status code was 404, but doesn't say which part it couldn't find. To debug I reduced the workflow down to 1 node with 2 cards. 1 gets the file, the other executes the code. Linking the 2 cards and errors below Is this the best way to let the client upload a file to the knowledgebase? And if so, how do I fix this error. Thank you https://cdn.discordapp.com/attachments/1366863664556150815/1366863664853942283/image.png?ex=68127e7c&is=68112cfc&hm=4cfbb669206b184613711e560480ef3035ed213eb2c462d5e0edf76bcbeb8549& https://cdn.discordapp.com/attachments/1366863664556150815/1366863665520709643/image.png?ex=68127e7c&is=68112cfc&hm=fe844b3edb3a0525899612317d625306ee767a39d146ca1ac9e3da9ff45b7b43& https://cdn.discordapp.com/attachments/1366863664556150815/1366863665810247760/image.png?ex=68127e7d&is=68112cfd&hm=8ad5a5cebd96bef918a3537db006093645b9513395c25138a5cae5a650706e5a& https://cdn.discordapp.com/attachments/1366863664556150815/1366863666107912313/image.png?ex=68127e7d&is=68112cfd&hm=0122d579b52c8622a1716de23ae8363960545e2c63eb9ba5bbbb4548fc10fb82&

Hi team, Can someone from the Botpress Cloud support team please help with the following request? Does Botpress Cloud use static IPs for outbound traffic? If so, could you please share the official list of static IP addresses (preferably in CIDR format)? Also, do you guarantee that these IPs won’t change without prior notice? Happy to reach out via the support form or email if needed,just let me know the best channel. Thanks in advance!

The chatbot is unable to utilize the knowledge base to respond to user questions. As a result, the chatbot simply stated that it does not have any knowledge concerning the questions when the information is clearly in the knowledge base

i published my bot and I've done the coding for it already but when i attempt to access it via the dashboard from webchat. it just keeps loading forever and never actually loads. Can you please help me?

When trying to access my Bot within my admin panel page (v12) I'm getting the following error message... 'Error occured while trying to proxy: botpressv12-19bpn.kinsta.app/studio/new-strollnsolve-march25' Any ideas? Thanks, Stevie

Hello i have trouble when i try to use custom CSS in bot setup for own styles for buble button -change icon, add some animation, size and so on. During load it show up for around 0,5s default styles, and after this it replace with my own styles from stylesheet. Is there someways to how load first custom stylesheet before default styles? Or atleas showup bubble icon after CSS are loaded completely? This is the part of code ****:root ..... * This i need to load first*

hi hi another quick question here i realize that conversation ID actually ties with the workflow timeout a lot so i am now using whatsapp cloud api to send first conversation, and i manually created a new conversation ID for every message i send out if, the phone number that previously has conversation ID with botpress did not time out, it kinda will go back to the latest conversation ID, and resume my conversation from there instead of following the new ID created is there any way that i can identify if my conversation ID has already time out? so that i can write the code to use back the previous conversation ID instead of creating a new one

Hi there, I was watching some tutorials and documentation on enabling HITL, including this one . However, in my case, there is no option for the HITL Agent under the Agents tab, even though I am subscribed to the Plus plan. Am I missing something? https://cdn.discordapp.com/attachments/1366699750191140904/1366699750375817257/image.png?ex=6811e5d4&is=68109454&hm=0e69bbfa4a82c9141974427d0616f142bd7a67d640d8c4b48076d1392d3feec0&

Hello, I was wondering if for the HITL function would be possible to integrate it with Crisp. And if there is no direct integration, if someone has a clue abou what I could do for integrate it. Thanks! https://cdn.discordapp.com/attachments/1366685457726963732/1366685457991209020/image.png?ex=6811d884&is=68108704&hm=4a0c7ef952beacf8887190b848d31747b37979ebd3c9bf9d6ec27ef48b15799a&

When i try to create a bot and get into the flows section, then creating a new note, the cards section is missing. Am i doing something wrong? (self hosted version) https://cdn.discordapp.com/attachments/1366633764226338896/1366633764737908798/Screenshot_2025-04-29_073136.png?ex=6811a860&is=681056e0&hm=c81e7fdcd7afe09d76befae0077bc700b4060f34083368cb7d6b54783499b472&

Hello, I am building a whatsapp bot for a restaurant that actively uses their whatsapp for business app on iphone. I have a new number that's specifically for the chatbot. Can I add this new number, request API access, and integrate with botpress without affecting the other numbers in the clients business whatsapp app?

Hello. The current goal is for the user to be able to upload a PDF and the bot extract information from the PDF. It was not working as "Vision" appeared to not be enabled. We were advised that you needed to upgrade to the Plus plan to enable vision, so we did. While we can enable "Vision" now, it appears the bot still cannot extract any information from uploaded files I tried uploading a PDF directly, converting the PDF to a jpg myself and uploading the jpg, as well as uploading the jpg to an image hosting site and giving the bot a link to the jpg. In all situations the inspect shows the bot cannot extract any content from the image. For relevance: -the bot is an Autonomous node -the "Vision Agent" is enabled -the autonomous node has the "Extract Content from Image" card -Vision is turned on in that area on the home screen under "Knowledgebase"

what are these [ nums ] in the responses? are they citining resources used? if so not doing properly, how can i turn this off? please let me know, thank you https://cdn.discordapp.com/attachments/1366490381579063296/1366490382111608934/image.png?ex=681122d7&is=680fd157&hm=fca66756575604407c81064bef2541c5d41989541d5757a9bc5b9d6d142fb9c4&

Hello everyone, I’m currently using Botpress Cloud and I’m facing an issue when trying to connect to an external service (Brevo) via API. Brevo has a security setting that only allows API calls from whitelisted IP addresses. However, it seems that the outbound IP address used by Botpress Cloud is dynamic or changes over time, which causes my API requests to fail when the IP is not recognized. • Is it possible to have a fixed/static IP for Botpress Cloud API calls? • Or is there an official way to handle this kind of situation? Thanks a lot in advance for your help or any advice!

Using the latest WebChat inject script, I’m attempting to set a custom userId during initialization. I’ve tested multiple inject versions and init() methods, but the userId continues to default or fail to apply. Current code is below. Would appreciate any input on what might be wrong. window.botpress.init({ "botId": "xxxxxxxxxxxxxxxxx", "userId": "gary d", "configuration": { "composerPlaceholder": "Chat", "botName": "Test", "botDescription": "Test Description", "website": {}, "email": {}, "phone": {}, "termsOfService": {}, "privacyPolicy": {}, "color": "#5eb1ef", "variant": "soft", "themeMode": "light", "fontFamily": "inter", "radius": 1 }, "clientId": "xxxxxxxxxxxxxx" });

I found a couple outdated threads with dead links. Does anyone have any information about what option exist to embed the bot not in a popup chatbot style but on a page directly with a search field ideally not via iframe.

Hi everyone, is it possible to translate the text here ? https://cdn.discordapp.com/attachments/1366433497123917855/1366433497484623944/image.png?ex=6810eddc&is=680f9c5c&hm=75c43437bd7e46ff7ac96b47de1940bf91a1cd5ddd7321819d730f6e83255ca8&

How to configure environment variables so that switching the bot from dev to production wouldn’t require major changes. Or the only way to do this is create two different chatbot for the dev and prod?

Hello team, there seems to be an issue with the {{conversation.SummaryAgent.transcript}} from summary agent. the transcript max lines translate to literal lines and not conversation, when we say line 10, it looks for 10 instances of "\n" within the transcript instead of the last 10 messages between the user and the bot. In my scenario here the bot is returning the following conversation.SummaryAgent.transcript = " - Pharmacy 1 \n [Location](https://maps.google.com/xxx) \n Phone: xxx\n - Pharmacy 2 \n [Location](https://maps.google.com/xxx) \n Phone:xxx\n\n**Sorry, I couldn't find that information.**\nbot: Would you like me to have someone from our team reach out to you via email regarding your unanswered query?\nuser: yes\n" https://cdn.discordapp.com/attachments/1366376827576385619/1366376827928449034/image.png?ex=6810b915&is=680f6795&hm=ce31ed366729516881ecbdfedf12994e9e78be74ccaf4d4e08cdbb2cb21a8275&

If I upload internal company documents to the Knowledge Base, will this data remain local within my Botpress environment, or is any of it shared externally (e.g., with Botpress servers)? Also, is any of this uploaded data used or intended to be used for training models within Botpress services, or does it stay strictly within my instance?

Hi Matthew, Patrick mentioned you should be able to help with this. We have been trying to make our integration public, but keep getting the red message "Integration does not meet the requirements to be made public, check documentation for more detail". We have followed all guidelines in the documentation. Do you mind assisting us in making it public? We have been trying for a few weeks now actually. We are not sure what it is missing! Thank you!

Hello, I am currently developing a lead-qualification WhatsApp chatbot, and since the system prompt/instructions are getting increasingly long, complex, and costly, I've been thinking of using multiple "agents"/autonomous nodes (ANs) instead of letting only one AN handle all the tasks. To start, I'd like to "delegate" the user-classification task to a separate AN. Specifically, I'd like the first AN to make a call to the second one with the necessary classification criteria. Then I'd like the second AN to return the group where the user falls into, according to the provided criteria. Then the first AN should proceed with the necessary steps, based on the group. My question is, how do I ideally go about this?. I've tried using the "Generate Content" card, but it's quite limited for more complex tasks, in particular, because it cannot have its own tools. I've also tried using a transition to another AN, but then the first AN doesn't really wait for a response, and I'm not sure how to send the second AN's response back to the first one in this case. Thanks for any hints!

Hello, is it possible to force the Autonomous Node to query the knowledge base before each of its answers of its conversation with the user?

Is there any way to limit the number of requests on a user device within a selectable time period? I want to avoid the risk of users spamming my chatbot credits. 🥺

Under "Knowledge bases" in my bot's "Home" section it list vision as disabled. When clicking it, it does not allow me to enable "Text (basic)" and under it "Vision" says upgrade to enable. When going to upgrade none of the packages mention vision, and I cannot find in the documentation (nor the AI search thing) what is required to actually enable vision. is it simply the 89/mo package? or is there an add-on for it?

I'm an extremely new user to botpress. I'm making a bot where the user uploads a PDF, then the bot extracts information and tells the user something based on it. While trying to get this set up I was notified i'm nearing the "File Storage" limit. It looks like any files uploaded by the user are stored for 48 hours. At no point will the bot need to access a file from a previous session. Is there a way to either have the bot not store uploaded files, or have it delete any uploaded files when the conversation ends? As I'm using an "Autonomous Node" I don't know that the workflow ever hits the "End" node, so I don't see an easy way to have a workflow trigger when "end" happens, nor do I see an easy way to have it get the ID of any files to use the "delete file" api

Hello Everyone, I am trying to send information via code execute to an email. For whatever reason, I can link my gmail account to botpress. Is there another way, I can send emails via botpress