Are API testing UIs really relevant anymore? I haven't manually tested an API for almost two years, and my APIs are SO much better. This kind of GUI is going to die over the next 5 years.
Postman should be seriously worried.
Postman should be seriously worried.
We have a postman collection mainly for testers to manually verify certain aspect of a otherwise fairly unit and end-to-end test covered API.
But sometimes the tester needs to make a test where they create something via the official API to use it in other parts of the app. They just open Postman (soon to be replaced), select the collection, select the server, select POST-item and of the the races. Not everything needs to be AI.
We could vibe code something specific for our needs, but that would be more or less again just end up something like postman again? At least I currently don't have the vision to see a big change.
This.
Why do you have a collection that is passed to humans for manual testing in the age of agents?
I find the CPU/disk space dashboards useless. Stuff like monit is useful but status of literal OS services is only part of the picture. There are many layers here and I sometimes wonder how others do this. Surely not all companies just have disposable fleets of “serverless” things?
...and... it's gone
Full disclosure, I'm building a desktop app called DB Pro [1] which is a modern desktop database client.
We're still doing well, and selling a lot, but we're having to think about what the next 2/3/4/5 years look like...
Trading bits you own for rented redundant compute to recreate same seems... suboptimal?
How much power should we burn on everyone's agent recreating curl?*
* yes, agents have tools, but what is and isn't an agent tool is a gray area. "Desktop software" could be an agent tool with a human UI.