Cindy Open-Source AI Agent Platform: Features, Multi-Device Capabilities and Licensing Overview

Overview of Cindy AI Agent

Cindy is an open-source AI agent platform designed to work out of the box across multiple devices, including desktop and mobile environments. According to project documentation, the platform allows users to connect different computers and phones to maintain continuous work sessions, access remote files, and share model providers without manual copying.

Key Functional Features

• Multi-Device Session Continuity: Users can open sessions from alternative devices, access project files, and continue conversations without re-establishing context.
• Model Sharing and Remote Desktop: The platform supports configured model providers from other connected computers, API providers, and local models. A built-in remote desktop feature allows phone users to operate connected computers when chat interfaces are insufficient.
• Agent Orchestration and Automation: Users can combine external harnesses like Claude Code, Codex, and Pi under a shared interface. The system supports multi-agent teams for parallel implementation and review, alongside routine automations and integrations via Slack, Telegram, Skills, and MCP.
• Rich Interactive Outputs: Responses can include charts, interactive controls, and HTML structures rather than text-only replies.

Technical Architecture and Repository Structure

The project client is organized as a pnpm monorepo containing open-source desktop and mobile clients alongside shared packages, licensed under the Apache License, Version 2.0. The backend service is maintained in a separate repository. Official distribution builds connect by default to official cloud services, while local operation is available via a local agent mode accessed through a skip sign-in option.

Analytics, Privacy, and Data Governance

Official distribution builds integrate TapDB for product-level aggregate usage statistics, tracking metrics such as device identifiers, OS versions, and startup events. According to the documentation:

• Consent Controls: Users can disable usage statistics through settings menus on desktop and mobile clients.
• Data Restrictions: Chat contents, file contents, and working directory data are not collected by TapDB.
• Local Mode Exemption: Running the app in local mode (skipping sign-in) does not report telemetry data.

Verification and Contribution

Developers can audit code, contribute via pull requests, and review complete architectural documentation within the repository's `docs/` folder and `CONTRIBUTING.en.md` guidelines. Regular contributions require a Developer Certificate of Origin sign-off (`git commit -s`).