aiconfig-update

To update LaunchDarkly AI Configs, fetch the current configuration first, then modify variation instructions, messages, models, or metadata parameters, and verify the results with a post-change API fetch.

The safest way to delete an AI Config is to first archive it reversibly, verify the archival state, and then execute the permanent delete flow with explicit safety checks to prevent accidental data loss.

Yes, you can archive LaunchDarkly AI Configs reversibly across projects instead of deleting them, allowing you to remove deprecated variations from active use while retaining the ability to restore them later.

Yes, you need a valid LaunchDarkly API token to authenticate requests when updating, archiving, or deleting AI Configs and their variations across different projects.

Fetching an AI Config before making changes enforces a safe governance workflow by verifying current state and metadata, preventing conflicting updates to variation instructions or models.

Yes, updating AI Config parameters supports post-change verification by executing an API fetch after the modification to validate that the new models, instructions, and metadata applied correctly.