主要更新
- 【重要更新】新增智能体模式（实验性），AI 可以读取屏幕并执行自动操作
- 新增模型列表配置页，支持从服务器获取模型列表
- 支持自定义功能按钮布局
- 思考内容支持折叠显示
- 新增数据导入导出功能
- 优化首次使用欢迎流程
- 支持设置上下文长度限制
- 支持单次回复并行调用多个 Tool，修复 Tool 调用时的上下文拼接问题
- 修复 gpt-5 系列导致的空指针和超时问题
Major Updates
- [Important Update] Newly added agent mode (experimental), where AI can read the screen and perform automatic operations
- Added a model list configuration page, supporting the retrieval of model lists from the server
- Support for customizable function button layout
- Support folding display for thinking content
- Added data import and export functionality
- Optimize the welcome process for first-time users
- Support setting context length limits
- Support parallel invocation of multiple Tools with a single response, and fix the context concatenation issue during Tool invocation
- Fixed null pointer and timeout issues caused by the GPT-5 series