
cindy
Consider it done. The open-source AI agent that works out of the box · 想到,就能做到。开源、开箱即用的 AI Agent。
Sign in to rate this tool.
Overview

Cindy is an open‑source AI agent that runs as soon as you install it. It lets you link multiple computers and a phone so you can keep a single conversation, task list, and model configuration across all devices. The platform is positioned as a ready‑to‑use solution that does not require you to deploy a backend server.
The tool is aimed at developers and power users who already have access to large language model subscriptions or local runtimes. You can attach Claude, CodeX, Pi, or other supported models by providing an API key or connecting a local runtime. Once linked, a model configured on one machine can be called from another, so you do not need to repeat setup on each device. The remote‑desktop feature shows the actual screen of a connected computer on your phone, letting you type and operate software remotely.
Cindy organizes work into tasks that retain context, memory, and a history of model usage. Within a task you can assign different “uses” – the execution engines tied to specific models, to separate agents. The interface supports multi‑agent workflows: you can plan a task, run parallel workers with different models, and then hand the result to a reviewer agent. Files attached to a task are browsable and transferable between devices, but the system does not perform whole‑drive synchronization. Connections to external services such as Slack, GitHub, or Feishu are available through installable plugins, and you can add custom skills or extensions to shape the agent's behavior.
Cindy is distributed via GitHub with builds for desktop and mobile platforms. The project is open source, so you can modify the code or add new integrations. Model usage is billed by the underlying provider; Cindy itself does not charge a subscription, but you must supply your own API keys or local hardware. Remote access must be enabled for each provider, and any data transferred to cloud models follows the provider's limits and fees.
Key features
- Connect multiple computers and a phone to a single Cindy instance
- Switch between Claude, CodeX, Pi, and other supported models within the same task
- Built‑in remote desktop lets you view and control a connected computer from mobile
- Multi‑agent workflow with task memory, history, and role‑specific uses
- Plugin system for Slack, GitHub, Feishu and custom skills
Details
- Pricing
- Free
- Category
- Code
Suggest an edit
Sign in to suggest an edit.
Pros and cons
Pros
- No separate backend server to deploy or rent
- Supports both cloud API keys and local model runtimes
- Allows model sharing across devices without reconfiguration
- Remote desktop integration works directly from the phone
Cons
- Pricing information is not published; costs depend on external model providers
- Local model support is listed as beta and may require specific hardware
- File access is limited to task‑specific transfers, not full drive sync
- Some third‑party tools need separate authorization, which is not detailed on the site

