- Introduced `CreateAgentV2Screen` and `CreateAgentV2ViewModel` for a new agent creation experience.
- Implemented AI-powered agent info generation based on user input, including title and description.
- Added a "manual mode" for users who prefer to input agent details directly.
- Enhanced UI with gradient borders, loading animations, and improved layout.
- Integrated avatar selection and cropping using `AgentImageCropScreen`.
- Refactored `AddAgentViewModel` to support state persistence across page navigation and to store generated input text.
- Updated API client to include a longer timeout for agent info generation requests.
- Added new drawable resources for UI elements.
- Switched `Const.BASE_SERVER` to use the release URL for debug builds.
- Replaced the old `AddAgentScreen` with the new `CreateAgentV2Screen` in navigation.
AllenTom
cf25540417
