makecindy/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。

What it is

Cindy is an open-source AI agent that runs on your own computer and does real work in your projects and apps. It's designed to work "out of the box" — ready on day one, but customizable over time.

Core concept

Cindy combines multiple harnesses (the underlying agent engines), models, and tools into a single agent. The first supported harnesses are Claude Code and Codex, with more planned and a native harness in development. Models and harnesses can be mixed freely and switched mid-task while your workspace, memory, skills, and tools stay continuous. A single task can even be planned, executed in parallel, and reviewed by agents running on different harness × model combinations. Cindy can also drive your browser, computer, and phone, and take work from instant messages and schedules.

What's in this repo

This repository is the open-source client for Cindy — the desktop and mobile apps plus their shared packages, organized as a pnpm monorepo:

Path Description
apps/desktop Electron desktop client
apps/mobile Expo / React Native mobile client
packages/* Shared client capabilities (auth, device-link, agent orchestration, model providers, …)
apps/*-bin Tool binaries shipped with the desktop app (downloaded per platform, not committed)

The backend service lives in a separate repository and is not part of this monorepo.

Models & accounts

You can bring models your way, with several options:

  • Sign in to the official Cindy service — usage deducted transparently
  • Authorize the Claude Code / Codex Coding Plan you already pay for and keep using it inside Cindy — no duplicate bill
  • Connect your own API keys
  • Use local models

There are two usage modes:

Mode Account requirement Availability
Hosted service Cindy cloud account Full hosted service, see pricing at cindy.app
Skip Sign-In No Cindy sign-in required Local agents only; server-backed capabilities unavailable

Customization ("Yours to shape")

  • Memory — correct her once and she does it right from then on, shared across harnesses
  • Skills — teach a way of working once and reuse it everywhere; team sharing is in the making
  • Automation — recurring work schedules itself, runs itself, reports back
  • MCP — wire your internal tools and business systems into her reach
  • Plugins — reshape features, UI, and interactions, shared through an open marketplace (in the making)
  • Source — audit, fork, extend, and contribute improvements back under Apache-2.0

Privacy & telemetry

Official distribution builds include TapDB usage analytics for aggregate statistics (device identifiers, OS/app version, startup and activity events). Events are associated with your account ID after sign-in, so it's not fully anonymous. TapDB does not collect chat content, file content, or working-directory data.

You can disable usage analytics without editing code: Desktop → Settings → About → Usage statistics; Mobile → Usage statistics in Settings. The preference defaults to on.

Other points:

  • Desktop stops reporting immediately when disabled; local mode (Skip sign-in) doesn't report to TapDB at all
  • Mobile immediately unlinks the account and stops explicit reports when disabled
  • Online heartbeats and diagnostic logs are separate mechanisms, unaffected by the usage statistics switch
  • Desktop native crash dumps stay local and are never uploaded automatically
  • Building from source: desktop dev builds don't initialize TapDB by default; mobile doesn't initialize without injected TapDB configuration

Getting started (development)

Prerequisites: Node.js 22.x, pnpm 10.x (v11 not yet supported), Git LFS

Minimal entry point:

git clone https://github.com/makecindy/cindy.git
cd cindy
git lfs pull
pnpm install

Development entry points:

pnpm restart:desktop:remote --region=cn      # Mainland China Cindy account
pnpm restart:desktop:remote --region=global  # Global Cindy account

Remote development uses your own Cindy cloud account and existing login state. The client connects to Cindy's official cloud services by default — external developers don't need to self-host a server.

Contributing & license

  • Contributions go through pull requests into main
  • Every commit needs a Developer Certificate of Origin sign-off (git commit -s); a DCO check enforces it, and no CLA is required
  • Code is licensed under Apache-2.0 (repository-root LICENSE governs; individual files don't carry headers)
  • Model weights, datasets, prompts, trademarks, and other separately identified materials may have their own license terms
  • Report security issues privately per SECURITY.en.md — never open a public issue

More info

  • Website: cindy.app · Downloads: cindy.app/download
  • Linux users: see the Ubuntu, Arch Linux, and Omarchy installation guide in docs/linux.md
  • Architecture docs: DESIGN.md, docs/README.md, docs/auth-realm-routing.md, AGENTS.md, docs/dev-rules/

Related

  • Project
    lahfir/agent-desktop

    agent-desktop is a cross‑platform CLI and C library that lets AI agents read the desktop UI via accessibility trees and control apps (click, type, window management, etc.) using JSON references.

  • Project
    agentty-org/agentty

    agentty is a native, single‑binary terminal coding assistant written in C++26. It talks to any LLM (Claude, GPT, Ollama, etc.), retrieves only the relevant parts of your repository via a hybrid BM25 + embedding RAG engine, runs shell/build commands inside OS sandboxes, and offers an optional Smart Mode that routes easy tasks to cheap models and complex planning to a flagship model. The tool starts in under 1 ms, has zero runtime dependencies, works on Linux/macOS/Windows/Android, and can operate in an air‑gapped SSH‑relay mode.

  • Project
    kody-hq/kody

    Kody is a Cloudflare Workers‑based platform that provides persistent memory, credential storage, code packages, and job scheduling for AI agents via the Model Context Protocol (MCP). It offers a Remix UI, OAuth‑protected endpoints, and a four‑worker architecture (origin, platform, runtime, jobs) to serve individual or team accounts. The repo is a full‑stack TypeScript monorepo (Nx) with local dev support, extensive docs, and a functional‑source license.

  • Project
    Gentleman-Programming/gentle-shell

    gentle‑shell is a terminal‑based coding‑agent built on the Pi LLM platform. It lets you describe a task in natural language, orchestrates the work through a structured “Organic Driven Development” workflow, can delegate to specialised sub‑agents, shows live Git diffs and agent status, and requires you to approve the exact change before committing. Installers exist for macOS, Windows and Linux, and the UI is customisable via a command palette and a `/gentle:customize` panel.

  • Project
    LodyAI/Lody

    Lody is an open‑source, cross‑platform workspace that lets teams share and coordinate AI coding agents (Claude Code, Codex, Kimi, etc.) across desktop, mobile, web and CLI. It syncs conversations, Git worktrees, files and visual feedback using CRDTs, enabling collaborative, version‑controlled AI‑driven development.