Repository navigation
Conversation
#968 made a bare `lk agent simulate` print help and exit 0. The dashboard's empty state, its Add to CI workflow, and scripts written before the text subcommand existed all run the bare form, so they silently did nothing (and passed in CI). A bare simulate runs text mode again; text and audio stay as explicit subcommands.
Contributor
Author
|
Before (v2.19.0): the dashboard's bare-simulate.mp4 |
Contributor
Author
|
Closing in favor of the decision on #1043: keep the explicit |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Since #968 (v2.18.7), a bare
lk agent simulateprints help and exits 0. That is the exact command the LiveKit Cloud dashboard tells new users to run on the empty Simulations page (lk agent simulate -n 5). It is also what the dashboard's "Add to CI" workflow runs, and what scripts written before thetextsubcommand existed run. On v2.18.7+ all of them silently do nothing, and in CI the step passes without running a single simulation.This restores the pre-#968 behavior: with no subcommand,
simulateruns text mode.text,audio,list,view,export,getandcancelare unchanged. The project still loads per subcommand, soget/list/cancelkeep skipping the API-key project gate.Impact: of projects that copied the empty-state command, the share that opened a simulation run within 7 days fell from 12.8% (Aug 15 – Sep 15, 33/258) to 4.5% (Sep 16 – Oct 15, 7/154). Part of that drop predates the release (web#5932 changed who sees the command), but every copy since v2.18.7 has been a no-op.
Tested:
go test ./cmd/lk/passes. Runninglk agent simulate -n 2 --yesfrom agent-starter-python on this branch generated scenarios and completed a run; on v2.19.0 it prints help.