让智能体，把事情推进到结果
开源的桌面与 CLI 全栈 AI 智能体。让目标、Files、工具和可检查的产出，留在同一项工作里。
写代码、查资料、处理Files和制作内容，都需要从理解问题走到实际产出。Lime 把项目上下文与工具执行连接起来，让Task可以逐步推进。
Lime 从一个目标start，阅读必要材料，在授权范围内修改Files、运行命令和调用工具。你可以检查过程、补充要求，再从已有进度continue。
项目Files与记录帮助你掌握工作进度。调用远程模型或外部工具时，相关内容会发送到你选择的服务；模型、工具和执行边界都应当由你明确选择。
Why is it designed this way?
Because creation is not a question and answer session. Data needs to be organized, drafts need to be revised, and the results need to be reviewed and reused. Lime wants to put those moves back into the same project.
Give a clear task entry first
After opening it, you first know what it can do, and then decide how to do it. You don't need to face a whole page of abstract abilities first.
Make the project a continuation object
Drafts, versions, release notes, and next steps should stay in the project, not scattered in chat logs.
Keep data local as much as possible
Reference materials, semi-finished products and sensitive context should be kept in the user's own environment as much as possible, with clearer boundaries.
We hope that the official website will present a clearer product expression: after opening it, you will first see the tasks, information and results, instead of being surrounded by a bunch of abstract concepts first.
So the Lime you see now will emphasize fewer and clear entrances, a larger product interface, and directly understandable usage paths. It should look like a real product, not a technology roadmap.
hello@lime.ai
business cooperation
business@lime.ai
Contact us
当前为演示表单，submit不会发送消息或保存个人资料。