docs: refresh v0.2 subscription copy (#323)

## Summary

<!-- One-paragraph description of the change. Focus on the *why*. -->

## Type of change

- [ ] Bug fix
- [ ] New feature
- [ ] Refactor (no behavior change)
- [ ] Documentation
- [ ] Build / CI / tooling
- [ ] Breaking change

## Linked issue

<!-- Use "Closes #123" only when this PR fully resolves the issue. Use
"Refs #123" for partial work or follow-up slices. -->

## Checklist

- [ ] I checked the linked issue / relevant context before starting
- [ ] `pnpm lint && pnpm typecheck && pnpm test` passes locally
- [ ] Added/updated tests for the change
- [ ] Added a changeset (`pnpm changeset`) if user-visible
- [ ] Updated docs if behavior changed

## Dependency additions (if any)

<!-- For each new prod dependency: name, install size, license,
why-not-alternatives. Delete this section if no new deps. -->

## Screenshots / recordings (UI changes)

<!-- Drop them here. Required for any visual change. -->

Co-authored-by: Sun-sunshine06 <Sun-sunshine06@users.noreply.github.com>

Author: Sun-sunshine06

.changeset/v02-subscription-copy.md

@@ -0,0 +1,7 @@
+---
+"@open-codesign/desktop": patch
+"@open-codesign/providers": patch
+"@open-codesign/i18n": patch
+---
+
+Refresh v0.2 user-facing copy for ChatGPT subscription sign-in and current provider support.

README.md

@@ -50,7 +50,7 @@
 
 Turn a prompt into a polished prototype, slide deck, or marketing asset, locally, with the model you already use.
 
-**Open CoDesign is the open-source Claude Design alternative** — built for people who want the speed of AI-native design tools without subscription lock-in, cloud-only workflows, or being forced onto a single provider. An MIT-licensed desktop app, local-first from day one, with BYOK for any model (Claude, GPT, Gemini, DeepSeek, Kimi, GLM, Ollama, or any OpenAI-compatible endpoint). One-click import of your existing Claude Code or Codex API key gets you running in under 90 seconds.
+**Open CoDesign is the open-source Claude Design alternative** — built for people who want the speed of AI-native design tools without subscription lock-in, cloud-only workflows, or being forced onto a single provider. An MIT-licensed desktop app, local-first from day one, with BYOK for any model (Claude, GPT, Gemini, DeepSeek, Kimi, GLM, Ollama, or any OpenAI-compatible endpoint) plus direct ChatGPT Plus / Pro / Team subscription sign-in for Codex models. One-click import of existing Claude Code or Codex provider configs, or one-click ChatGPT sign-in, gets you running in under 90 seconds.
 
 ---
 
@@ -85,7 +85,7 @@ Open source, desktop-native, and built for people who do not want their design w
 | Version history | ✅ Local sessions + workspace files | ❌ | ❌ | ❌ |
 | Data privacy | ✅ On-device app state | ❌ Cloud-processed | ❌ Cloud | ❌ Cloud |
 | Editable export | ✅ HTML, PDF, PPTX, ZIP, Markdown | ⚠️ Limited | ⚠️ Limited | ⚠️ Limited |
-| Price | ✅ Free app, token cost only | 💳 Subscription | 💳 Subscription | 💳 Subscription |
+| Price | ✅ Free app, provider/subscription cost only | 💳 Subscription | 💳 Subscription | 💳 Subscription |
 
 ---
 
@@ -128,7 +128,7 @@ Open source, desktop-native, and built for people who do not want their design w
 
 **Time to first artifact:** about 3 minutes
 
-**Requires:** one API key or local Ollama
+**Requires:** one API key, ChatGPT subscription sign-in, or local Ollama
 
 **Runs on:** macOS 12+ (Monterey or later), Windows 10+, Linux (glibc ≥ 2.31)
 
@@ -184,16 +184,15 @@ After each stable tag push, CI syncs SHAs back into `packaging/` and publishes d
 >
 > Want a verified build? Compile from source — see [CONTRIBUTING.md](./CONTRIBUTING.md).
 
-### 2. Add your API key
+### 2. Add a provider
 
-On first launch, Open CoDesign opens the Settings page. Paste any supported provider key:
+On first launch, Open CoDesign opens the Settings page. Pick the path that matches how you already use models:
 
-- Anthropic (`sk-ant-…`)
-- OpenAI (`sk-…`)
-- Google Gemini
-- Any OpenAI-compatible relay (OpenRouter, SiliconFlow, local Ollama)
+- **ChatGPT subscription** — sign in with ChatGPT to use Codex models without pasting an API key.
+- **API key** — paste Anthropic (`sk-ant-...`), OpenAI (`sk-...`), Google Gemini, OpenRouter, SiliconFlow, DeepSeek, or another supported provider key.
+- **Local / keyless** — use Ollama or an IP-allowlisted OpenAI-compatible gateway.
 
-Credentials stay in `~/.config/open-codesign/config.toml` (file mode 0600, same convention as Claude Code, Codex, and `gh` CLI). Nothing leaves your machine unless your chosen model provider requires it.
+Credentials stay in `~/.config/open-codesign/config.toml` and the ChatGPT OAuth token store under the app config directory. Nothing leaves your machine unless your chosen model route requires it.
 
 ### 3. Type your first prompt
 
@@ -203,7 +202,7 @@ Pick one of **fifteen built-in demos** — landing page, dashboard, pitch slide,
 
 ## Bring your stack
 
-Already using Claude Code or Codex? Your providers, models, and API keys import in one click, with no copy-paste and no need to re-enter settings:
+Already using Claude Code or Codex? API-key provider configs import in one click, with no copy-paste and no need to re-enter settings. If you use Codex through ChatGPT subscription login, sign in directly from Settings:
 
 ![Import from Claude Code or Codex in one click](https://raw.githubusercontent.com/OpenCoworkAI/open-codesign/main/website/public/demos/claude-code-import.gif)
 
@@ -223,14 +222,14 @@ Add a `SKILL.md` to any project to teach the model your own taste.
 
 ### Models and providers
 - **Unified provider model** — Anthropic, OpenAI, Gemini, DeepSeek, OpenRouter, SiliconFlow, local Ollama, or any OpenAI-compatible relay; keyless (IP-allowlisted) proxies supported
-- **One-click import** from Claude Code and Codex configs — bring your existing providers, models, and keys in a single click
+- **One-click import and sign-in** — bring Claude Code / Codex API-key provider configs across, or sign in with ChatGPT subscription for Codex models
 - **Dynamic model picker** — every provider exposes its real model catalogue, not a hardcoded shortlist
 
 ### Generation and editing
 - **Prompt → HTML or JSX/React component** prototype, rendered in a sandboxed iframe (vendored React 18 + Babel on-device)
 - **Fifteen built-in demos + twelve design skill modules** — ready-to-edit starting points for common design briefs
 - **Live agent panel** — watch tool calls stream in real time as the model edits files
-- **AI image generation** — opt-in bitmap assets for heroes, product shots, backgrounds, and illustrations
+- **AI image generation** — opt-in bitmap assets for heroes, product shots, backgrounds, and illustrations via OpenAI, OpenRouter, or signed-in ChatGPT subscription
 - **AI-generated sliders** — the model emits the parameters worth tweaking (color, spacing, font)
 - **Comment mode** — click any element in the preview to drop a pin, leave a note, and let the model rewrite only that region
 - **Generation cancellation** — stop mid-stream without losing prior turns
@@ -303,7 +302,7 @@ Have a different priority in mind? [Open an issue](https://github.com/OpenCowork
 
 - Electron + React 19 + Vite 6 + Tailwind v4
 - `@mariozechner/pi-ai` and `pi-coding-agent` (model/provider and agent-loop primitives)
-- `better-sqlite3`, `electron-builder`
+- `electron-builder`
 
 ## Reporting issues
 

README.zh-CN.md

@@ -48,7 +48,7 @@
 
 Open CoDesign 可以把一句自然语言提示词，直接变成一个完成度很高的 HTML 原型、幻灯片或者营销素材，而且整个过程都可以在你的电脑上完成。
 
-它适合这样一类人：想要 AI 设计工具的速度，但不想被订阅制绑住，不想把工作流全丢到云端，也不想只能用某一家模型。你可以把它理解成一个更开放、更本地化的 Claude Design 替代方案：开源、桌面原生、支持自带 API Key，也支持多模型切换。
+它适合这样一类人：想要 AI 设计工具的速度，但不想被订阅制绑住，不想把工作流全丢到云端，也不想只能用某一家模型。你可以把它理解成一个更开放、更本地化的 Claude Design 替代方案：开源、桌面原生、支持自带 API Key、ChatGPT 订阅登录和多模型切换。
 
 ---
 
@@ -83,7 +83,7 @@ Open CoDesign 可以把一句自然语言提示词，直接变成一个完成度
 | 版本历史 | ✅ 本地 session + 工作区文件 | ❌ | ❌ | ❌ |
 | 数据隐私 | ✅ 应用状态保留在本地 | ❌ 云端处理 | ❌ 云端 | ❌ 云端 |
 | 可编辑导出 | ✅ HTML、PDF、PPTX、ZIP、Markdown | ⚠️ 有限制 | ⚠️ 有限制 | ⚠️ 有限制 |
-| 价格 | ✅ 应用免费，仅承担模型 token 成本 | 💳 订阅制 | 💳 订阅制 | 💳 订阅制 |
+| 价格 | ✅ 应用免费，仅承担 provider 或订阅成本 | 💳 订阅制 | 💳 订阅制 | 💳 订阅制 |
 
 ---
 
@@ -126,7 +126,7 @@ Open CoDesign 可以把一句自然语言提示词，直接变成一个完成度
 
 **第一次生成一个结果，大概只要：** 3 分钟
 
-**你需要准备：** 一个 API Key，或者本地 Ollama
+**你需要准备：** 一个 API Key、ChatGPT 订阅登录，或者本地 Ollama
 
 **支持平台：** macOS、Windows、Linux
 
@@ -182,16 +182,15 @@ scoop install opencoworkai/open-codesign
 >
 > 想要可验证构建可以自己从源码编译，见 [CONTRIBUTING.md](./CONTRIBUTING.md)。
 
-### 2. 添加 API Key
+### 2. 添加 provider
 
-首次启动时，Open CoDesign 会直接打开设置页。你可以粘贴任意支持的 provider key：
+首次启动时，Open CoDesign 会直接打开设置页。按你已有的模型入口选择即可：
 
-- Anthropic（`sk-ant-…`）
-- OpenAI（`sk-…`）
-- Google Gemini
-- 任意 OpenAI 兼容中继，比如 OpenRouter、SiliconFlow、本地 Ollama
+- **ChatGPT 订阅登录**：直接登录 ChatGPT，使用 Codex 模型，无需粘贴 API Key。
+- **API Key**：粘贴 Anthropic（`sk-ant-...`）、OpenAI（`sk-...`）、Google Gemini、OpenRouter、SiliconFlow、DeepSeek 或其他支持的 provider key。
+- **本地 / Keyless**：使用 Ollama，或使用 IP 白名单的 OpenAI 兼容网关。
 
-凭证会保存在 `~/.config/open-codesign/config.toml`（文件权限 0600，与 Claude Code、Codex、`gh` CLI 的做法一致）。除非你选择的模型提供商本身需要联网，请求内容不会额外离开你的机器。
+凭证会保存在 `~/.config/open-codesign/config.toml`，ChatGPT OAuth token 会保存在应用配置目录下的 token store 中。除非你选择的模型入口本身需要联网，请求内容不会额外离开你的机器。
 
 ### 3. 输入第一条提示词
 
@@ -201,7 +200,7 @@ scoop install opencoworkai/open-codesign
 
 ## 可以直接接你现有的开发环境
 
-如果你已经在用 Claude Code 或 Codex，现有的 provider、model 和 API key 都可以一键导入，不用复制粘贴，也不用重新配一遍。
+如果你已经在用 Claude Code 或 Codex，API key 形式的 provider、model 和 key 都可以一键导入；如果你用的是 Codex 的 ChatGPT 订阅登录，可以直接在设置里登录 ChatGPT。
 
 ![一键导入 Claude Code 或 Codex 配置](https://raw.githubusercontent.com/OpenCoworkAI/open-codesign/main/website/public/demos/claude-code-import.gif)
 
@@ -221,14 +220,14 @@ scoop install opencoworkai/open-codesign
 
 ### 模型与提供商
 - **统一的 provider 抽象**：支持 Anthropic、OpenAI、Gemini、DeepSeek、OpenRouter、SiliconFlow、本地 Ollama，以及任意 OpenAI-compatible relay；同时支持无 key 的 IP 白名单代理
-- **一键导入 Claude Code 和 Codex 配置**：现有的 provider、model 和 key 可以直接带进来
+- **一键导入和登录**：Claude Code / Codex 的 API key provider 配置可以直接带进来，也可以用 ChatGPT 订阅登录使用 Codex 模型
 - **动态模型选择器**：每个 provider 都会展示真实模型列表，而不是一小撮写死的选项
 
 ### 生成与编辑
 - **提示词 → HTML 或 JSX/React 组件原型**，渲染在隔离的 sandbox iframe 中（本地 vendored React 18 + Babel）
 - **十五个内置 demo + 十二个设计技能模块**：给常见设计需求准备好的起点
 - **实时 Agent 面板**：模型编辑文件时，工具调用会实时流式展示
-- **AI 图像生成**：可选启用，为 hero、产品图、背景和插画生成位图素材
+- **AI 图像生成**：可选启用，可通过 OpenAI、OpenRouter 或已登录的 ChatGPT 订阅为 hero、产品图、背景和插画生成位图素材
 - **AI 自动生成调节参数**：模型会主动暴露值得调的参数，比如颜色、间距和字体
 - **Comment mode**：点击预览中的任意元素，留下批注，模型只重写对应局部
 - **支持中途取消生成**：停止后也不会丢失之前的上下文和结果
@@ -293,7 +292,7 @@ v0.2 已经把 Open CoDesign 从一次性生成器升级成一个本地设计 ag
 
 - Electron + React 19 + Vite 6 + Tailwind v4
 - `@mariozechner/pi-ai` 与 `pi-coding-agent`（模型 / provider 与 agent loop 基础能力）
-- `better-sqlite3`、`electron-builder`
+- `electron-builder`
 
 ## 社群
 

apps/desktop/src/main/onboarding-ipc.test.ts

@@ -2,8 +2,8 @@
  * Tests for settings IPC channel versioning.
  *
  * These tests verify that registerOnboardingIpc exposes only the versioned
- * v1 settings channels. v0.2 has not shipped yet, so renderer/main IPC drift
- * should fail loudly instead of being hidden by unversioned compatibility
+ * v1 settings channels. v0.2 is the shipped IPC surface, so renderer/main IPC
+ * drift should fail loudly instead of being hidden by unversioned compatibility
  * handlers.
  */
 
@@ -778,6 +778,32 @@ describe('getApiKeyForProvider — API key retrieval', () => {
 });
 
 describe('config:v1:import-codex-config empty env handling', () => {
+  it('guides Codex ChatGPT subscription users to the built-in ChatGPT sign-in path', async () => {
+    await registerIpcForTest();
+    const { readCodexConfig, codexAuthPath } = await import('./imports/codex-config');
+    const { mkdir, writeFile } = await import('node:fs/promises');
+    const { tmpdir } = await import('node:os');
+    const { join } = await import('node:path');
+    vi.mocked(readCodexConfig).mockResolvedValueOnce({
+      providers: [],
+      activeProvider: null,
+      activeModel: null,
+      envKeyMap: {},
+      apiKeyMap: {},
+      warnings: [],
+    });
+    const dir = join(tmpdir(), `codesign-codex-chatgpt-${Date.now()}-${Math.random()}`);
+    await mkdir(dir, { recursive: true });
+    const path = join(dir, 'auth.json');
+    await writeFile(path, JSON.stringify({ auth_mode: 'chatgpt' }), 'utf8');
+    vi.mocked(codexAuthPath).mockReturnValueOnce(path);
+
+    const handler = handlers.get('config:v1:import-codex-config');
+    await expect(handler?.({} as unknown)).rejects.toThrow(
+      /Open CoDesign now supports ChatGPT subscription directly/,
+    );
+  });
+
   it('rejects imports that only found an empty env credential', async () => {
     const { readCodexConfig } = await import('./imports/codex-config');
     const { encryptSecret } = await import('./keychain');

apps/desktop/src/main/onboarding/external-imports.ts

@@ -64,7 +64,7 @@ export async function runImportCodex(imported: CodexImport): Promise<OnboardingS
   if (imported.providers.length === 0) {
     throw new CodesignError(
       (await detectChatgptSubscription())
-        ? 'Detected Codex ChatGPT subscription login (auth_mode: chatgpt). It cannot be imported as an API-key provider yet — the "Sign in with ChatGPT subscription" feature is still being polished and will ship in the next release. For now, configure [model_providers] in ~/.codex/config.toml manually, or switch to API-key mode in Codex. / 检测到 Codex 使用 ChatGPT 订阅登录，无法自动导入为 API key provider。"用 ChatGPT 订阅登录"功能仍在打磨中，下个版本开放 —— 目前请在 ~/.codex/config.toml 里手动配置 [model_providers]，或改用 API key 登录 Codex。'
+        ? 'Detected Codex ChatGPT subscription login (auth_mode: chatgpt). Open CoDesign now supports ChatGPT subscription directly, but Codex config import only reads API-key [model_providers] from ~/.codex/config.toml. Open Settings > Models and use "Sign in with ChatGPT subscription" to add the ChatGPT provider. / 检测到 Codex 使用 ChatGPT 订阅登录。Open CoDesign 现在已支持直接使用 ChatGPT 订阅，但“从 Codex 导入”只读取 ~/.codex/config.toml 中的 API key [model_providers]。请到 设置 > Models 点击“用 ChatGPT 订阅登录”添加 ChatGPT 订阅 provider。'
         : 'No importable API provider found in Codex config (~/.codex/config.toml is missing a [model_providers] section). / Codex 配置里没有可导入的 API provider（~/.codex/config.toml 里缺少 [model_providers] 段）。',
       ERROR_CODES.CONFIG_MISSING,
     );

package.json

@@ -2,7 +2,7 @@
   "name": "open-codesign",
   "version": "0.2.0",
   "private": true,
-  "description": "Open-source AI design tool — prompt to interactive prototype, slide deck, and marketing assets. Multi-model, BYOK, runs on your laptop.",
+  "description": "Open-source AI design tool — prompt to interactive prototype, slide deck, and marketing assets. Multi-model, BYOK or ChatGPT login, runs on your laptop.",
   "license": "MIT",
   "homepage": "https://github.com/OpenCoworkAI/open-codesign",
   "repository": {

packages/i18n/src/locales/en.json

@@ -16,7 +16,7 @@
     "comingSoon": "Coming soon",
     "loading": "Loading…",
     "preAlpha": "pre-alpha",
-    "tagline": "BYOK · local-first · multi-model",
+    "tagline": "BYOK · ChatGPT login · local-first",
     "done": "Done.",
     "applied": "Applied.",
     "working": "Working"
@@ -268,7 +268,7 @@
     },
     "imageGen": {
       "tabTitle": "Image generation",
-      "tabHint": "Let the design agent call an image model (e.g. gpt-image-2) for heroes, posters, and background bitmaps. Credentials stay local.",
+      "tabHint": "Let the design agent call an image model through OpenAI, OpenRouter, or ChatGPT subscription for heroes, posters, and background bitmaps. Credentials stay local.",
       "title": "Image generation assist",
       "hint": "When enabled, the agent can invoke generate_image_asset during a run.",
       "enabled": "Enable image generation assist",
@@ -628,7 +628,7 @@
         "429": "Rate limited. Wait a moment and try again.",
         "network": "Cannot reach base URL. Check domain/port/network.",
         "unsupported": "Unrecognized key prefix. Supported: sk-ant- (Anthropic), sk- (OpenAI), sk-or- (OpenRouter).",
-        "notSupportedProvider": "{{provider}} is not supported in v0.1. Use Anthropic, OpenAI, or OpenRouter.",
+        "notSupportedProvider": "{{provider}} is not supported by the first-run provider shortcut. Use Anthropic, OpenAI, OpenRouter, Ollama, ChatGPT subscription sign-in, or add a custom provider in Settings.",
         "rendererDisconnected": "Renderer is not connected to the main process.",
         "detectIpc": "Provider detection failed (main process unreachable): {{message}}. Restart the app and try again.",
         "detectNetwork": "Provider detection failed (network error): {{message}}. Check your connection and try again."

packages/i18n/src/locales/es.json

@@ -16,7 +16,7 @@
     "comingSoon": "Próximamente",
     "loading": "Cargando…",
     "preAlpha": "pre-alpha",
-    "tagline": "BYOK · local-first · multi-modelo",
+    "tagline": "BYOK · inicio con ChatGPT · local-first",
     "done": "Listo.",
     "applied": "Aplicado.",
     "working": "Trabajando"
@@ -239,7 +239,7 @@
     },
     "imageGen": {
       "tabTitle": "Generación de imágenes",
-      "tabHint": "Permite que el agente de diseño llame a un modelo de imagen (ej. gpt-image-2) para encabezados, pósteres y mapas de bits de fondo. Las credenciales permanecen locales.",
+      "tabHint": "Permite que el agente de diseño llame a un modelo de imagen mediante OpenAI, OpenRouter o una suscripción de ChatGPT para encabezados, pósteres y mapas de bits de fondo. Las credenciales permanecen locales.",
       "title": "Asistente de generación de imágenes",
       "hint": "Cuando está habilitado, el agente puede invocar generate_image_asset durante una ejecución.",
       "enabled": "Habilitar asistente de generación de imágenes",
@@ -573,7 +573,7 @@
         "429": "Límite de tasa excedido. Espera un momento e intenta de nuevo.",
         "network": "No se puede alcanzar la URL base. Revisa dominio/puerto/red.",
         "unsupported": "Prefijo de clave no reconocido. Soportados: sk-ant- (Anthropic), sk- (OpenAI), sk-or- (OpenRouter).",
-        "notSupportedProvider": "{{provider}} no es compatible en la v0.1. Usa Anthropic, OpenAI o OpenRouter.",
+        "notSupportedProvider": "{{provider}} no es compatible con el acceso rápido del primer inicio. Usa Anthropic, OpenAI, OpenRouter, Ollama, inicia sesión con una suscripción de ChatGPT o añade un proveedor personalizado en Ajustes.",
         "rendererDisconnected": "El renderizador no está conectado al proceso principal.",
         "detectIpc": "Error al detectar proveedor (proceso principal inalcanzable): {{message}}. Reinicia la app e intenta de nuevo.",
         "detectNetwork": "Error al detectar proveedor (error de red): {{message}}. Revisa tu conexión e intenta de nuevo."