llmx.txt and llms-full.txt can't be synced with HubSpot Customer Agent

[michaelkremmel]

@kneth suggested we add https://cdn.crate.io/about/v1/llms.txt and https://cdn.crate.io/about/v1/llms-full.txt as a knowledge source to the HubSpot Customer Agent (AI-powered Goatie Chatbot on our website).

Unfortunately, those URLs couldn't be synced and HubSpot doesn't provide any reason why the sync failed.

So it might be just a HubSpot issue, but I wanted to let you know anyway. It isn't a crucial error for the Customer Agent training, as it can still sync data from the docs regardless of those two files, and I could also upload the files as a knowledge source (rather than sync them from the URLs, but in this case, they are static and not automatically updated.)

FYI On the topic of how crucial llms.txt and llms-full.txt are, I got the following feedback from HubSpot recently:

According to the team, there's no edge on being an early adopter with llms.txt for most of our customers. It's been a lot of hype and talk without actual physical evidence that LLMS are visiting and using llms.txt. Additionally, as of mid-July, our team re-did our check of web traffic going to HubSpot-hosted pages and it confirmed we're not seeing evidence that AI systems are using llms.txt.

They also provided this link where John Mueller, Google Search Advocate, is cited with "FWIW no AI system currently uses llms.txt."

[amotl]

Hi and thanks. I also don't have any answers, just links.

References

• https://ref.tools/
• https://www.mintlify.com/blog/how-often-do-llms-visit-llms-txt
• https://medium.com/google-cloud/give-your-ai-agents-deep-understanding-with-llms-txt-4f948590332b

Examples

• https://github.com/bytebase/dbhub/blob/main/llms-full.txt
• https://langchain-ai.github.io/langgraph/llms.txt

[amotl]

Observations

FYI On the topic of how crucial llms.txt and llms-full.txt are, I got the following feedback from HubSpot recently:

According to the team, there's no edge on being an early adopter with llms.txt for most of our customers. It's been a lot of hype and talk without actual physical evidence that LLMS are visiting and using llms.txt. Additionally, as of mid-July, our team re-did our check of web traffic going to HubSpot-hosted pages and it confirmed we're not seeing evidence that AI systems are using llms.txt.

They also provided this link where John Mueller, Google Search Advocate, is cited with "FWIW no AI system currently uses llms.txt."

Interesting.

• https://www.seokratie.de/llms-txt/
• https://de.semrush.com/blog/llms-txt/
• https://www.sistrix.de/frag-sistrix/ai-grundlagen/llms-txt/

It looks like SEO companies and consultants became technological advisors today, many blowing into the same horn?

Die Verwendung von llms.txt lohnt sich im Moment wahrscheinlich nicht.

Evaluation I

[there is no] actual physical evidence that LLMS are visiting and using llms.txt

Well. How would they? What's going on here again?

Evaluation II

John Mueller, Google Search Advocate, is cited with "FWIW no AI system currently uses llms.txt."

Is it another anti-democratization campaign by GAFAMs & Co.?