Skip to main content
Glama
README.md
# Codex with ChatGPT

> ChatGPT thinks. Codex works.
> ChatGPT 负责思考,Codex 负责干活。

## The problem · 解决什么问题

**中文** — ChatGPT 付费订阅的网页版额度大量闲置,Codex 却在消耗紧张的
API 额度做规划和 Review。本项目把"思考"交给你已付费的网页版 ChatGPT,
Codex 只负责执行。不用 API Key、不搞逆向代理——官方网页 + 只读 MCP 桥接。

**EN** — ChatGPT Plus/Pro web quota sits idle while your coding agent burns
scarce API/Codex tokens on planning and review. This project moves the
thinking to the subscription you already pay for; Codex only executes.
No API keys, no reverse proxy — official web UI plus a read-only MCP bridge.

## What it is · 这是什么

**中文** — 把 ChatGPT 网页版变成 Codex 编码会话的"规划与审查大脑",执行权
完全保留在 Codex 手里。你的仓库永远不会被上传:ChatGPT 通过一条安全的、
OAuth 保护的**只读** MCP 连接,按需读取当前工作区里它真正需要的那几行代码。

**EN** — Use the ChatGPT web app as the planning and review brain for your
Codex coding sessions, while Codex keeps full ownership of execution. Your
repository is never uploaded: ChatGPT reads exactly the lines it needs through
a secure, OAuth-protected, **read-only** MCP connection to your current
workspace.

Detailed docs below are in English · 详细中文文档见 **[README.zh-CN.md](README.zh-CN.md)**

## One-paste install · 一段话安装

**中文** — 不懂 git、Node、终端?完全不需要懂。把下面这段话原样复制给你的
编码 Agent(Codex),然后去倒杯咖啡:

```text
请帮我完整安装并配置 Codex with ChatGPT,全程自动,我是不懂技术的小白,
所有事情你自己做:

1. 环境自检:需要 git 和 Node.js ≥ 20,缺什么就自动安装
  (macOS 用 Homebrew,Windows 用 winget),同时安装 cloudflared。
2. 下载:把 https://github.com/XiaoDuoYa/codex-with-chatgpt 克隆到
   ~/codex-with-chatgpt(已存在就 git pull 更新)。
3. 构建:在该目录里执行 corepack pnpm install 和 corepack pnpm build。
4. 安装 Skill:把仓库里的 skill/SKILL.md 复制到
   ~/.codex/skills/codex-with-chatgpt/SKILL.md,并把文件中
   "The codex-with-chatgpt checkout lives at:" 那一行的路径改成实际克隆路径。
5. 首次配置:按 SKILL.md 里的 first-time setup 流程执行
  (运行 c2c setup,用内置浏览器打开 ChatGPT 配置连接器并输入配对码)。
   全程只用内置浏览器,禁止打开任何第三方浏览器。
6. 只有遇到需要我登录(ChatGPT / Cloudflare)、验证码或两步验证时才叫我,
   而且一次只告诉我一个动作。
7. 完成后给我看 ✓ 清单,并确认文件读取测试通过。我不懂 MCP、OAuth、
   Tunnel、端口这些词,不要向我解释;出了问题先自己修。
```


**EN** — Don't know git, Node, or terminals? You don't need to. Copy the
paragraph below, paste it to your coding agent (Codex), and go grab a coffee:

```text
Please install and configure "Codex with ChatGPT" for me, fully automatically.
I am a non-technical user — do everything yourself:

1. Check the environment: git and Node.js >= 20 must be available. Install
   anything missing yourself (macOS: Homebrew, Windows: winget). Also install
   cloudflared.
2. Download: clone https://github.com/czstudio/codex-with-chatgpt-pro into
   ~/codex-with-chatgpt (if it already exists, git pull to update).
3. Build: inside that folder run `corepack pnpm install` then `corepack pnpm build`.
4. Install the Skill: copy skill/SKILL.md to
   ~/.codex/skills/codex-with-chatgpt/SKILL.md, and update the line
   "The codex-with-chatgpt checkout lives at:" to the actual clone path.
5. First-time setup: follow the SKILL.md "first-time setup" workflow
   (run c2c setup, configure the ChatGPT connector in the BUILT-IN browser,
   enter the pairing code). Never open a third-party browser.
6. Only interrupt me for logins (ChatGPT / Cloudflare), CAPTCHAs or 2FA —
   and give me exactly ONE action at a time.
7. When done, show me the ✓ checklist and confirm the file-read test passed.
   I don't know what MCP, OAuth, tunnels or ports are. Don't explain them.
   If anything breaks, fix it yourself first.
```


**Updates · 更新** — The Skill checks GitHub once a day and updates itself when a
new version is released; no action needed. You can also say "更新 Codex with ChatGPT"
anytime. / Skill 每天自动检查一次 GitHub,有新版本会自动更新,无需任何操作;
也可以随时对 Codex 说"更新 Codex with ChatGPT"。

---

*The sections below are in English. 以下详细内容为英文,中文完整版见
[README.zh-CN.md](README.zh-CN.md)。*

## Install → Setup → Use (manual)

### One-command installer

macOS (Homebrew installed):

```bash
curl -fsSL https://raw.githubusercontent.com/czstudio/codex-with-chatgpt-pro/main/scripts/install.sh | bash
```

Windows 10/11 (PowerShell; uses winget):

```powershell
irm https://raw.githubusercontent.com/czstudio/codex-with-chatgpt-pro/main/scripts/install.ps1 | iex
```

The installer checks prerequisites, builds the bridge, installs the Codex Skill,
and configures its persistent app-state directory. It never uploads a repository
or stores ChatGPT/Cloudflare credentials in the project.

1. Install the Codex Skill: copy `skill/` to `~/.codex/skills/codex-with-chatgpt/`.
2. Tell Codex: **"Set up Codex with ChatGPT."** (中文: "使用 Codex with ChatGPT 完成首次配置。")
3. Use Codex normally: **"Use Codex with ChatGPT to implement XXX."**

That's the whole manual. You don't need to know what MCP, OAuth, tunnels,
ports or localhost are — Codex configures everything automatically and you
just see:

```
Codex with ChatGPT

✓ Project detected
✓ Workspace Bridge started
✓ Secure connection established
✓ ChatGPT connected
✓ File read test passed

Ready.
```

The only step that may need you: logging into ChatGPT (and, if you want a
stable hostname, logging into Cloudflare once).

### Optional stable hostname

The default public address is a temporary Cloudflare URL. It changes when the
bridge restarts, and Codex repairs ChatGPT by deleting that workspace's
connector and adding it again.

If you have a Cloudflare account and a domain already on Cloudflare, first-time
setup (and the next coding session, once) will ask whether you want a stable
hostname such as `c2c-<project>.your-domain.com`. That path opens a browser so
you can authorize Cloudflare. After that, the ChatGPT connector keeps working
across restarts. If you skip it, or login fails, Codex stays on the temporary
address — same features, just a slower repair.

Credentials stay in the OS app state directory, not in the project.

## How it works

```
             ┌───────────────────────────┐
             │       ChatGPT Web         │
             │  Reason / Plan / Review   │
             └──────────┬──────────▲─────┘
                        │          │
               MCP      │          │ Computer Use
            Data Plane  │          │ Control Plane (<1 KB messages)
                        ▼          │
             ┌─────────────────────┐
             │      C2C Bridge     │   loopback-only HTTP server
             │  read-only MCP      │   OAuth 2.1 + one-time pairing code
             │  OAuth + Pairing    │   Cloudflare Quick Tunnel
             │  Tunnel Manager     │
             └──────────┬──────────┘
                        │  read-only
                        ▼
             ┌─────────────────────┐          ┌─────────────────────┐
             │   Local Workspace   │◀─────────│    Codex Harness    │
             └─────────────────────┘ edit/git │ shell / tests / fix │
                                              └─────────────────────┘
```

- **Control plane (Computer Use)**: Codex and ChatGPT exchange tiny structured
  `[C2C]` state messages — `INIT → PLAN → EXECUTED → REVIEW → DONE`. No diffs,
  no logs, no file bodies are ever pasted.
- **Data plane (MCP)**: ChatGPT pulls what it needs itself through 8 read-only
  tools: `workspace_info`, `list_directory`, `read_file`, `search_workspace`,
  `git_status`, `git_diff`, `test_status`, `execution_summary`.
- **Independent review**: after Codex executes, ChatGPT inspects the actual
  git diff and test records through MCP — it never trusts "all tests passed"
  claims blindly.
- **Recoverable task cursor**: compact task-bound checkpoints preserve the
  current state, known issues, and next step across compaction or chat handoff.
- **Fail-closed evidence gate**: C2C cannot claim DONE without matching execution,
  tests, and review checkpoint evidence. It returns non-zero when incomplete and
  also blocks on malformed, conflicting, unknown-schema, or oversized audit data.
  It remains local protocol enforcement—not authoritative project state.

## Security model (short version)

- **Read-only by construction**: write/delete/shell/commit tools simply do not
  exist on the server. No prompt injection can enable them.
- **One workspace = one boundary**: every token is bound to a single workspace;
  path containment uses canonical realpaths (symlink/`../`/absolute-path escapes
  are all blocked and tested).
- **Sensitive files never leave**: `.env*`, keys, SSH, credentials are denied by
  default (`.env.example` allowed); `.c2cignore` adds your own rules.
- **Knowing the URL grants nothing**: the public MCP endpoint requires OAuth 2.1
  (PKCE S256, dynamic client registration, rotating refresh tokens). Without a
  token: 401. Wrong workspace: 403.
- **The model never sees long-lived credentials**: the only secret that ever
  touches a browser is a one-time pairing code (5-minute TTL, 5 attempts,
  rate-limited, destroyed on use).

Full threat model: [docs/security.md](docs/security.md)

## For developers

```bash
pnpm install
pnpm build          # -> dist/, exposes the `c2c` bin
pnpm test           # path security, OAuth, pairing, recovery and MCP e2e

c2c setup           # bridge + tunnel + pairing code, all in one
c2c sandbox-allow   # whitelist the settings dir in Codex (macOS + Windows)
c2c status / doctor / pair / unpair / logs / stop
```

Requirements: Node.js >= 20, git. `cloudflared` for the public connection
(auto-detected; the Skill installs it for you).

Docs: [architecture](docs/architecture.md) · [protocol](docs/protocol.md) ·
[security](docs/security.md) · [troubleshooting](docs/troubleshooting.md)

## Project layout

```
src/
  bridge/     loopback HTTP server, port recovery, admin API
  mcp/        8 read-only tools, stateless Streamable HTTP
  auth/       OAuth 2.1 (PKCE, DCR, refresh rotation, revocation)
  pairing/    one-time pairing codes (CSPRNG, TTL, rate limits)
  workspace/  path containment, sensitive-file policy, search, git
  tunnel/     TunnelProvider abstraction + Cloudflare Quick/Named Tunnel
  execution/  execution records for the review loop
  process/    daemon lifecycle
  cli/        the c2c CLI
skill/        the Codex Skill (the real UX layer)
tests/        unit + integration tests
docs/         architecture / protocol / security / troubleshooting
```

## Status & disclaimer

V1. Verified end-to-end: bridge, OAuth + pairing, public tunnel, ChatGPT
connector setup, zero-touch first-run experience.

**Unofficial community project. Not affiliated with or endorsed by OpenAI.**

## License

[MIT](LICENSE)