This is an early release preview. You may encounter bugs.
Cindy logo
Unclaimed

Tool coding interface

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。

A 85/100 GitHub score ? This grade is derived from GitHub signals, not user votes. Open for the full breakdown.
Discussion

01 / Summary

About Cindy.

Cindy is a desktop and mobile AI agent that combines several agent harnesses, models, and tools into one assistant for work in your projects and apps. It runs locally on your machine against your real files and logged-in apps, and it can drive your browser, computer, and phone, and take work from instant messaging and schedules. This repository holds the open-source client: an Electron desktop app, an Expo and React Native mobile app, and their shared packages. The backend service is in a separate repository.

Claude Code and Codex are the first supported harnesses, with Pi listed on the project site and a native harness in development. Models and harnesses can be mixed and switched mid-task while the workspace, memory, skills, and tools stay continuous. A single task can be planned, executed in parallel, and reviewed by agents running on different harness and model combinations.

Models can come from the hosted Cindy service, from an existing Claude Code or Codex coding plan authorized inside the app, from your own API keys, or from local models. A Skip Sign-In mode runs local agents without a Cindy account, with server-backed features unavailable. Memory and skills are shared across harnesses, recurring work can be scheduled, MCP connects internal tools, and a plugin system is in development.

  • Multiple harnesses: Claude Code and Codex today, with more harnesses and a native one in development
  • Mid-task switching: change harness or model during a task while workspace, memory, skills, and tools carry over
  • Multi-agent tasks: plan, run parallel workers, and review with different harness and model combinations
  • Local execution: works on your own machine with your real files and logged-in apps
  • Device control: drives the browser, the computer, and a phone
  • Flexible model access: hosted service, an existing coding plan, your own API keys, or local models
  • Shared memory and skills: corrections and taught workflows persist across harnesses
  • Automation: recurring work runs on a schedule and reports back, and tasks can arrive from instant messaging
  • MCP: connects internal tools and business systems
  • Desktop and mobile clients: an Electron desktop app and a React Native mobile app

02 / Discussion CREDIBILITY-GATED

Discussion

Reading is open to everyone. Posting and voting need a verified identity or a GitHub grade of B or higher.

  • No discussions yet.

04 / Build

Build with Cindy.

Browse the catalog for frameworks, tools, and harnesses, each scored on real GitHub credibility.

Get Cindy →

Browse the catalog