You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Since we now can generate KQL via OpenAI (#577), the inverse could be super useful as well.
Adding a button or another way to send the current KQL query to OpenAI and get back an explanation in plain language could be really useful.
Mockup:
Example prompt:
"You are a KQL expert and a Grafana expert that can explain any KQL query that contains Grafana macros clearly to someone who isn't familiar with KQL or Grafana. Explain the following KQL. \nKQL: DestinyActivities | where $__timeFilter(TimeGenerated) | order by TimeGenerated asc"
The response back looks something like this:
"content": "This KQL query is selecting data from the DestinyActivities table, which likely contains information about activities within the video game Destiny. \n\nThe `where $__timeFilter(TimeGenerated)` portion is using a Grafana macro to filter the results based on the time range selected in the Grafana dashboard. This ensures that only data from the specified time period is included in the query results.\n\nFinally, the `order by TimeGenerated asc` portion sorts the query results in ascending order based on the TimeGenerated field, which likely indicates when each activity occurred. \n\nOverall, this query is selecting and filtering data about Destiny activities and sorting it in chronological order based on the time it occurred."
Another aspect to solve is how to display the result, we could:
Inject the response as a comment into the editor directly
Inject the response above or below the editor and somehow draw attention to it
Show the result in a modal or pop-up of some kind
The text was updated successfully, but these errors were encountered:
Since we now can generate KQL via OpenAI (#577), the inverse could be super useful as well.
Adding a button or another way to send the current KQL query to OpenAI and get back an explanation in plain language could be really useful.
Mockup:
Example prompt:
The response back looks something like this:
Another aspect to solve is how to display the result, we could:
The text was updated successfully, but these errors were encountered: