hi there, i have remixed a Chat with webpage Chrome extension agent. it works fine for websites. however, i am having issue ‘‘reading’’ content within webpage of my learning mgt system - which has a PDF render view within the webpage (i am logged into the system with my acct). any idea what i can do? tk u.
Hi @el2030,
Thanks for the post!
It looks like the PDF is being displayed as a visual embed on the page, which means the extension can’t actually access or read the content inside it.
There are a couple of options:
- If possible, try switching the content to regular text instead of a PDF
- Or download the PDF and create a simple agent where you can upload the file and chat about it
thanks so much for reply Alex. enc is another page that my chrome extension can’t see to read. any idea what i can do?
Hi @el2030,
Looks like this might be the same situation as your earlier post. The content is likely inside an iframe or another kind of embedded container, which means the Chrome Extension can’t access it.
As a workaround, you can create an Agent with a User Input set to Long Text. That way, you can copy and paste the content and chat with the AI from there.
Thanks very much. based on my check,
It looks like the MindStudio Chrome extension can’t extract or access the page content from Brightspace’s d2l/le/enhancedSequenceViewer
because the actual learning content is dynamically loaded inside a custom web component (<d2l-sequence-viewer>
) via shadow DOM or JavaScript fetch calls.
- The HTML source I am seeing doesn’t contain the actual learning content.
- Brightspace loads content dynamically via APIs, so the raw HTML (
rawHtml
,pageContent
, etc.) your extension receives is just a shell.
Would like to check if it’s possible to
Update the MSr Chrome extension logic to:
- Wait for the full page render (e.g.,
setTimeout
orMutationObserver
). - Traverse the shadow DOM if needed (e.g.,
.shadowRoot.querySelector(...)
) from<d2l-sequence-viewer>
.
Tk u,