tedix-hq/tedix is a platform where people and AI agents handle shared work, project knowledge, apps, and approvals together. The cli-v0.19.3 release on GitHub changed how the team's executor and model selection are managed, using explicit path configuration. When there is no configuration, automatic assignment is held off, and OpenCode is required to have a model specified. The source of these changes is the cli-v0.19.3 release on GitHub.
This change is an approach to state clearly in configuration which executor and model a task is assigned to. According to the project description, tedix is a platform that coordinates people and AI agents through shared work, project knowledge, apps, and approvals. For such a use case, executor and model selection are components that make up agent work. This release includes behavior that does not assign automatically when the selection is not determined, and requires specifying the model directly in OpenCode.
However, the release information does not describe the specific format of the path configuration, nor the process that follows when automatic assignment is held off for lack of configuration. It also gives no information on whether existing configurations need to be changed or what impact this has on other behavior. The verifiable scope is therefore limited to this: executor and model selection become explicit configuration, and the specification rules for the two cases have changed. The actual configuration method and scope of application are hard to judge from this release information alone.