Internal tool using n8n for **test case generation** based on user stories/features stored in Azure DevOps. I want to simplify the workflow trigger for my team using a basic **form-like web interface**, ideally built directly in n8n

Questions
1

:wrench: What I want to achieve:

  1. User visits a URL (e.g., /webhook/test-case-generator)
  2. A form appears: “Enter Feature ID” ➜ [Submit]
  3. On submit:
    • n8n fetches the user story from Azure DevOps (description + acceptance criteria)
    • Sends data to Azure OpenAI (GPT-4)
    • Categorizes test cases into :white_check_mark: Positive, :cross_mark: Negative, :orange_square: Edge Cases
    • Posts results to Microsoft Teams
    • Optionally returns the result to the browser or as .txt file. how to achieve this
2

Which part would you like our help with?

  • I am sure you found the form node by now, right?
  • Fetching from Azure DevOps will require making an API call using http request node.
  • Categorizing test cases into categories - depending on what you mean by categorizing (this needs to be deduced or read as a property) - you will need to either sort through or use an AI node to do that
  • Finally, posting to teams - there is a dedicated node for that as well.
3

@jabbson i want to know how kw to achieve this entire workflow. Im very new to this n8n.

Currently am doing poc on test case generation with help of azure openai, azure board where we have userstories and features based on on description and acceptance criteria. I need to pass there id which need to fetch description and acceptance criteria. Once user type the id it need to fetch user stories or features from that it have to pass to azure AI to generate testcase either as a browser plugin or via teams. Is that possible?

You understand my workflow right?

If you provide me the json so i can try to recreate it

4

Hey there! I haven’t personally worked with Azure, so I may not be the best person to guide you through the specifics — but I’d still like to offer a bit of friendly advice that might help you get better support here.

From my experience (and there’s quite a bit of it), you’ll get much more useful help if you break things down into smaller, focused questions. Instead of saying “can someone do this whole thing for me,” it’s better to approach it like: “I tried X, but instead of Y, I’m getting Z - here’s what I did, and here are the outputs or screenshots.” That kind of detail really encourages people to jump in and help.

And don’t worry if you’re just getting started - we all were at some point! What really makes a difference is showing some initiative: try things out, do a bit of digging, and then ask for help when you hit a wall. With tools like AI assistants these days, you can often get quick answers without having to wade through tons of docs.

If you’re genuinely looking to build experience with n8n, it’s totally worth investing some time into learning it properly. There are great resources right on the n8n website - video tutorials, written guides, and a mountain of YouTube content out there.

Stick with it, and you’ll be surprised how much you can pick up quickly!