Replies: 1 comment 4 replies
-
I have found that the results from the tool are correct. For example, I loaded up my project with 790 teams. If I ask the tool to use 790, it returns them all. However, Copilot gets utterly confused. The output is all there, but it does not know how to just list them out. But if I take the results and put in ChatGPT, it works fine. I am going to move this to a discussion and close the Issue. I don't think we have a problem with the tool itself. Maybe other folks have some ideas. |
Beta Was this translation helpful? Give feedback.
4 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
I gave a query in the form of "Find FOO team in BAR project". It seems like the agent will first try to query with only "top = 200" and say that the team doesnt exist.
Problem 1: its output just assumes it doesn't exist. it doesn't have the context to know it only tried the top 200 teams
After insisting on it to query more than just 200 teams, it tries with "top = 1000". It still cannot find it.
Problem 2: With "top = 1000" it seems to only return 720 results. Does the ADO API endpoint have a limit of 720? if so the tool should probably account for that?
After insisting that "FOO" team definitely should exist and it should query more, it just continuously queries with "top = 1000" many times... all returning the same 720 results each time
Problem 3: The agent doesn't seem to understand the skip parameter of the tool at all and never attempts to us it.
After telling it its just giving me the same results and it isnt trying to use the skip parameter, it finally is able to use "skip" to correctly paginate and eventually finds team FOO in project BAR
Beta Was this translation helpful? Give feedback.
All reactions