cindy
SHA-256Open-source, out-of-the-box AI agent client unifying multiple harnesses and models with shared memory, skills, and automation.
Smart Download
Download Download Version
v0.1.115 · 260.5 MB
An open-source agent that orchestrates Claude Code, Codex, and more to finish real work on your machine.
Core Features
- Works out of the box—finishes real work in your projects and apps from day one
- Orchestrates multiple harnesses (Claude Code, Codex, more coming) with mid-task model switching
- Memory and skills shared across harnesses—teach once, reuse everywhere
- Built-in automation: recurring tasks schedule, run, and report back by themselves
- Drives your browser, computer, and phone; accepts tasks from IM and schedules
What It Can't Do
- •The backend service is closed-source and not in this repository—only the client is open, so full self-hosting is not possible. Official builds include TapDB analytics (aggregate device/OS/version metadata only, no chat or file content); build from source to remove it. "Skip Sign-In" mode disables all server-backed capabilities. Full cloud features require the paid Cindy service. Building from source strictly requires Node.js 22.x, pnpm 10.x (v11 unsupported), and Git LFS. Desktop auto-updates come from the official CDN.
Use Cases
- Delegating real coding and file work in local projects to an AI that operates your actual machine
- Scheduling recurring jobs like daily report generation that run and report automatically
- Reusing an existing Claude Code or Codex subscription inside one unified app without double-paying
- Having different harness × model combos plan, execute in parallel, and review the same complex task
- Wiring internal business systems into the agent via MCP
Cindy is an open-source, out-of-the-box AI agent whose desktop and mobile clients are Apache-2.0 licensed. It runs locally, using your real files and logged-in apps, and can drive your browser, computer, and phone. Compared with single-harness tools like Claude Code or Codex CLI, its key differentiator is cross-harness orchestration: harnesses and models mix freely, switch mid-task, and share continuous memory, skills, and tools—one task can be planned, executed in parallel, and reviewed by agents on different combos. It also offers scheduling, MCP integration, and plugins. Bring models your way: the official Cindy service, an existing Claude Code/Codex plan with no duplicate billing, your own API keys, or local models. The backend service is closed-source; official builds include TapDB analytics (no chat/file content) that can be removed in source builds.
Tags
Getting Started
Download installer
Click the button above to download the installer for your system
Install the software
Open the downloaded dmg file, then drag the app to Applications
Download the installer for your platform from cindy.app/download
Run the installer and follow the prompts; Linux users should follow the official Ubuntu/Arch guide
Launch the app and sign in to a Cindy account, or choose "Skip Sign-In" for local-only mode
- Download the installer for your platform from cindy.app/download
- Run the installer and follow the prompts; Linux users should follow the official Ubuntu/Arch guide
- Launch the app and sign in to a Cindy account, or choose "Skip Sign-In" for local-only mode
SHA-256 checksum verified
Checksum extracted from GitHub official Release page
SHA256 Checksum
b2e36f19461e8c3d5d596d37973e2762466d0cace42f1f0e17e4e7e3f56be31dThis checksum is extracted from the GitHub Release page. Verify file integrity after download.
All SHA-256 checksums on this platform are extracted from the project's official GitHub Release page, without any modification. You can independently verify them on the GitHub Releases page.
Open Source Transparency
View GitHub SourceUninstall Info
On Windows, uninstall via Settings > Apps; on macOS, drag the app to the Trash; on Linux, remove it with your package manager. For a complete cleanup, manually delete Cindy's configuration and data folders in your user directory, and remove any stored credentials from the system keyring.
No Extra Dependencies
Ready to use after download. No additional runtime required.
Having issues? Check the FAQ below
5 FAQs
Similar Projects
ollama
Ollama lets you download, run, and manage large language models locally. One command, multiple platforms, endless possibilities.
ComfyUI
ComfyUI is a modular AI engine with a node-based graph interface for generating images, video, 3D models, and audio. Unlike traditional webui tools, it gives pixel-level control over every model and parameter, making it the preferred choice for professionals.
llama.cpp
High-performance LLM inference engine in C/C++ with minimal dependencies, supporting quantized models (1.5–8 bit) and diverse hardware (Apple Silicon, CUDA, Vulkan, etc.).