Thanks — great question. You can absolutely manage environment-specific configs (like

staging.toml

and

prod.toml

) for LiveKit deployments, but the

lk

CLI expects a single

livekit.toml

file by name. Here are practical patterns you can use to keep separate config files per environment. • Use separate LiveKit Cloud projects for each environment (staging, prod). Each project has its own configuration and secrets. • Keep distinct files locally (for example,

staging.toml

and

prod.toml

) and copy or rename the one you need to

livekit.toml

before running the

lk

CLI: a.

cp staging.toml livekit.toml

(deploy staging) b.

cp prod.toml livekit.toml

(deploy prod) • Automate the swap in CI/CD (recommended): ◦ In GitHub Actions or your pipeline, choose the correct TOML based on branch, environment, or GitHub environment and copy it to

livekit.toml

before the deploy step. ◦ Store secrets/API keys in environment variables or use a secrets file (the CLI supports providing secrets separately) so you don’t embed sensitive values in the TOML. • If you prefer strict separation, maintain separate projects and pipeline credentials so deployments can run independently without manual renaming. If you’d like, I can: • Show an example GitHub Actions snippet that copies the right file and runs the

lk

deploy step. • Recommend a secrets-handling pattern for your pipeline. Which would be most helpful for you? Sources: Self-hosted deployments | LiveKit Documentation | Deploy LiveKit Agents with GitHub Actions | LiveKit | Google AI and LiveKit | LiveKit Documentation