We would like the HubSpot MCP server to support creating draft Knowledge Base articles directly in HubSpot.
Right now an AI assistant connected through the MCP server, in our case Claude, can generate the article content, but we have to copy and paste it into HubSpot manually.
The workflow we are after:
- The AI assistant generates the article content.
- It sends it to HubSpot through the MCP server as a draft Knowledge Base article.
- We review and publish it manually inside HubSpot.
Being able to create drafts would be enough for us. Publishing should stay a manual step in HubSpot, so the approval stays with us.
That would cut the manual work while keeping publishing approval inside HubSpot. It would be really helpful.
Keeping publication approval manual makes sense. A useful first step might be to automate just the handoff into a draft, then measure how much copy/ paste time it removes and what formatting still needs review. What kind of content structure or metadata would the draft need to preserve?
Thank you for the reply. We agree that the handoff into a draft is a good first step. We estimate it would save us about an hour per article and make a real difference to the quality of our knowledge base.
On structure: The draft should be as close to ready as possible, so the review is about the content and not about formatting. For us that means:
- Title and subtitle.
- The article body with headings, lists, links and basic formatting.
- Images and videos, if possible.
- Category and subcategory.
- Tags, so the article can be found in the knowledge base search.
Images, videos and category are the most important for us.
It would also make a big difference if the MCP server could read existing articles and save updates to them as drafts. Keeping the knowledge base up to date takes a lot of our time, and being able to review and update articles this way would help just as much as creating new ones. Publishing would still be a manual step in HubSpot.
Thank you for picking this up.