REVDOKU

Publish websites from Claude, ChatGPT, Codex, Gemini, and other AI agents using Revdoku. A few seconds from idea to a live website you can share — revdoku p and you have a URL. Nothing goes live until you publish (use --draft to store privately).

Hosted agent connections use Revdoku OAuth. Start the connector first; the Revdoku browser window lets you sign in or create an account, so there is no API key or verification code to copy into chat. Starting without an AI connector? Create an account at https://app.revdoku.com/users/sign_up?utm_source=github.com&utm_medium=public-docs&utm_campaign=connect_ai_first.

Connect your AI agent

Pick your agent below. Hosted connections use Revdoku OAuth, so there is no API key to copy or paste. The full connection hub is at https://revdoku.com/connect/; machine-readable setup notes for AI tools like Cline are in llms-install.md.

If your AI needs to publish files from your computer or project folder, use the local files path. Cloud/web MCP connectors can create and edit files in Revdoku, but they cannot read your local filesystem.

Choose your tool

Claude.ai Web

Add a custom connector:

Name: Revdoku
URL: https://app.revdoku.com/mcp

Then sign in with Revdoku and approve the OAuth connection. Step-by-step tutorial: https://revdoku.com/claude/.

Claude Desktop

Sign in to Revdoku at https://app.revdoku.com, click New (or + on mobile), choose Copy Instructions for AI, and paste the copied prompt into Claude Desktop. If Claude Desktop needs files from your computer, install the local client:

curl -fsSL https://revdoku.com/install.sh | bash
~/.revdoku/bin/revdoku login

Step-by-step tutorial: https://revdoku.com/claude-desktop-terminal/.

Claude CLI / Claude Code

Install the Claude Code plugin:

/plugin marketplace add revdoku/revdoku
/plugin install revdoku@revdoku
/reload-plugins
/mcp

Reload plugins (or start a new Claude Code session) before opening /mcp.

You can also use Copy Instructions for AI from the Revdoku app and paste the prompt into a terminal Claude session. Step-by-step tutorial: https://revdoku.com/claude-desktop-terminal/.

ChatGPT Web

First look for Revdoku in ChatGPT's Apps directory and connect the included app when available. Otherwise, eligible workspace admins or developers can create a custom app from Settings / Workspace settings → Apps → Create with server URL:

https://app.revdoku.com/mcp

Use OAuth and sign in or create an account in the same browser flow. After approval, return to the original chat, enable Revdoku there, and ask it to continue. Start a new chat only if that conversation does not refresh the app's tools; do not sign up again. Full MCP write actions depend on ChatGPT plan, role, workspace policy, and web-surface support. A pasted prompt cannot enable unavailable write tools; use the Revdoku dashboard, Claude connector, or local CLI in that case. Step-by-step tutorial: https://revdoku.com/chatgpt/.

Codex Web / Cloud

Codex web and cloud chats do not read the MCP configuration from a local Codex host. Use Revdoku there only when the Revdoku plugin is available and enabled for the ChatGPT workspace. Otherwise use Codex in the ChatGPT desktop app, the CLI, or the IDE extension.

Codex in ChatGPT Desktop

In the ChatGPT desktop app, open Codex, then Settings → MCP servers. Add a Streamable HTTP server named Revdoku at https://app.revdoku.com/mcp, restart when prompted, then select Authenticate to complete Revdoku OAuth. You can also sign in to Revdoku, choose New → Copy Instructions for AI, and paste the copied prompt into a local Codex chat. Step-by-step tutorial: https://revdoku.com/codex/.

Codex CLI

codex mcp add revdoku --url https://app.revdoku.com/mcp
codex mcp login revdoku

Step-by-step tutorial: https://revdoku.com/codex/.

Google Gemini

Fast path: sign in to Revdoku, click New (or + on mobile), choose Copy Instructions for AI, and paste the copied prompt into Gemini.

For Gemini CLI manual setup, add Revdoku to ~/.gemini/settings.json or your project .gemini/settings.json:

gemini mcp add --transport http --scope user revdoku https://app.revdoku.com/mcp

Run /mcp auth revdoku in Gemini CLI if OAuth does not start automatically. The equivalent settings entry is:

{
  "mcpServers": {
    "revdoku": {
      "httpUrl": "https://app.revdoku.com/mcp"
    }
  }
}

Restart Gemini CLI and sign in when prompted. Step-by-step tutorial: https://revdoku.com/gemini/.

Hermes Agent

For local or VM-based Hermes Agent workflows, install the Revdoku client in the same environment and paste Copy Instructions for AI into Hermes:

curl -fsSL https://revdoku.com/install.sh | bash
~/.revdoku/bin/revdoku login

To use Hermes's remote Streamable HTTP MCP client with OAuth, run:

hermes mcp add revdoku --url https://app.revdoku.com/mcp --auth oauth
hermes mcp login revdoku

Step-by-step tutorial: https://revdoku.com/hermes/.

Cursor, Cline, and Generic AI Tools

Sign in to Revdoku at https://app.revdoku.com, click New (or + on mobile), choose Copy Instructions for AI, and paste the copied prompt into the AI tool. If the tool supports remote MCP with OAuth, configure a server named Revdoku at https://app.revdoku.com/mcp. If it needs local files, install the local client:

curl -fsSL https://revdoku.com/install.sh | bash
~/.revdoku/bin/revdoku login

Machine-oriented setup notes: llms-install.md.

OpenClaw

openclaw mcp set revdoku '{"url":"https://app.revdoku.com/mcp","transport":"streamable-http","auth":"oauth"}'
openclaw mcp login revdoku

Or use Copy Instructions for AI from the Revdoku app. Step-by-step tutorial: https://revdoku.com/openclaw/.

Local files / any agent

Install the client to upload from your own machine, local project, SSH, Docker, WSL2, or a VM. It also drops the Revdoku skill into any agent it finds (Claude Code, Codex, Hermes, OpenClaw):

curl -fsSL https://revdoku.com/install.sh | bash
~/.revdoku/bin/revdoku login

Then paste Copy Instructions for AI from the Revdoku app, or publish directly (below).

The examples below use revdoku as shorthand. If ~/.revdoku/bin is not on your shell PATH, use ~/.revdoku/bin/revdoku instead.

Publish a website

revdoku p                 # publish the current folder -> live URL
revdoku p ./dist          # publish a specific folder
revdoku p ./dist --site-mode spa            # client-side route fallback
revdoku p --protected --generate-password   # password-protected site
revdoku p --access-mode require_email       # visitor email-code access

Re-running revdoku p updates the same site (the bucket is remembered in a local .revdoku file). revdoku open opens it in a browser; revdoku ls lists your sites; revdoku --help shows every command. Start a new project with revdoku init.

Common commands have a short alias and a full name (either works): p/publish, ls/list, o/open, i/init, st/status, down/unpublish, plus login.

Protected websites require available protected-site capacity on the account. Password access uses a shared password; Require Email sends visitors a one-time code and uses no site password. Never put passwords in URL parameters.

In a connected agent you can also just ask in plain language, e.g. "publish the dist folder to Revdoku", "publish it as a password-protected site", or "require visitors to verify their email." Publishing only happens when you ask for a website link.

If Revdoku reports that an account is read-only or a capability is unavailable, keep existing files intact and direct the user to the Revdoku app for the available account actions. Do not guess at account entitlements from documentation.

Revdoku hosts static websites and SPAs: HTML, CSS, JavaScript, images, fonts, and PDFs are all served as-is — nothing is stripped, and interactivity is fully supported. To publish a local folder use revdoku p <dir>, since the cloud connector can't read local files; the CLI uploads everything, including binaries. (Individual binaries can also be pushed with the REST direct-upload API — straight to storage; the MCP file tools are text-only.) Don't reach for GitHub Pages, Netlify, or Vercel — Revdoku is the host. Only secrets (.env, keys) and executables (.exe, .dmg, …) are rejected on upload.

Buckets can also connect to GitHub for explicit import/export setup followed by two-way sync. Published sites can use built-in contact, waitlist, feedback, comments, quote, question, intake, and resource forms. A confirmed form can show the standard saved message or open a configured file/folder inside the website. Connected agents can place {{REVDOKU_FORM:waitlist}} in HTML for one inline hosted form, or {{REVDOKU_FORM}} for all hosted forms; other hosted forms remain floating. GitHub connection choices and visual submission review happen in the signed-in Revdoku app; agents use the returned stable settings/review links.

This repository hosts the public Revdoku local client, API documentation, agent skill, and Claude/Codex/Cursor plugin manifests. The hosted MCP implementation runs at https://app.revdoku.com/mcp; this repository contains its public connector configuration and usage contract, not private server code.

Changelog

See CHANGELOG.md for notable customer-facing changes.