English Français العربية

Puppetflow

Official registry Search and web

Build, run, monitor, and control Puppetflow browser automations from AI clients.

Source code Website Documentation

What can you do with Puppetflow?

  • Build browser workflows in JavaScript
  • Run workflows manually, on a schedule, or through integrations
  • Watch browser sessions live and take control when needed
  • Organize flows and reusable snippets in workspaces
  • Store, tag, organize, and edit text files in a scoped Media Library, then reuse them in flow inputs or $upload
  • Manage variables, secrets, and browser profiles
  • Inspect run history, logs, and execution results
  • Connect repositories and external vaults like 1Password

Install

Add this to your client's MCP configuration (Claude Desktop: claude_desktop_config.json; Cursor: .cursor/mcp.json; VS Code and others use the same "mcpServers" shape).

Remote (streamable-http)
Claude Code 
claude mcp add --transport http puppetflow 'https://{instance_host}/api/workspaces/{workspace}/mcp-server/http'
JSON configuration (Claude Desktop, Cursor, VS Code) 
{
    "mcpServers": {
        "puppetflow": {
            "type": "http",
            "url": "https://{instance_host}/api/workspaces/{workspace}/mcp-server/http"
        }
    }
}

Documentation

From the project's README on GitHub, refreshed weekly. Links open on GitHub.

Contents
  1. Puppetflow
  2. 🌐 Website and documentation
  3. 🎉 Features
  4. 🔗 n8n integration
  5. 🤖 MCP server for AI clients
  6. Puppetflow in action
  7. 🛟 Discussion / Need help?
  8. Project philosophy
  9. A note on project history
  10. No AI Slop Policy
  11. Repository documentation
  12. License

Puppetflow

Puppetflow is a self-hosted browser automation platform for writing, running, scheduling, and monitoring JavaScript workflows. Watch browser sessions live and take control whenever a run needs human attention.

⭐️ Like Puppetflow? Give the repository a star. It helps others discover the project.

Install Puppetflow with one command

curl -fsSL https://puppetflow.com/install.sh | bash

🌐 Website and documentation

🎉 Features

  • Build browser workflows in JavaScript
  • Run workflows manually, on a schedule, or through integrations
  • Watch browser sessions live and take control when needed
  • Organize flows and reusable snippets in workspaces
  • Store, tag, organize, and edit text files in a scoped Media Library, then reuse them in flow inputs or $upload
  • Manage variables, secrets, and browser profiles
  • Inspect run history, logs, and execution results
  • Connect repositories and external vaults like 1Password
  • Integrate with messaging tools and mailboxes
  • Deploy on your own infrastructure with Docker
  • Extend Puppetflow with reusable flows and snippets
  • Automate Media Library files and folders through the authenticated REST API

🔗 n8n integration

Puppetflow provides the n8n-nodes-puppetflow community node for adding browser automation to n8n workflows.

  • Trigger a flow with optional JSON input.
  • List and monitor runs, then retrieve their results.
  • Download screenshots, files, and session recordings.

Install the package from Settings > Community Nodes on a self-hosted n8n instance, then connect it with your Puppetflow instance URL and an API key from Profile > API Keys. See the n8n integration guide for setup details.

🤖 MCP server for AI clients

Every Puppetflow instance exposes a remote Model Context Protocol server over Streamable HTTP. Connect Cline, Cursor, Claude, Claude Code, or any other MCP client to search, inspect, create, and run browser automation flows, follow runs, and retrieve results, screenshots, and recordings.

Nothing runs locally. Enable MCP under Workspace > Settings > Instance MCP, create an access token, then add the endpoint to your client:

{
  "mcpServers": {
    "puppetflow": {
      "type": "streamableHttp",
      "url": "https://your-puppetflow.example/api/mcp-server/http",
      "headers": {
        "Authorization": "Bearer mcp_your_token_here"
      }
    }
  }
}

An OAuth endpoint with PKCE and Dynamic Client Registration is also available for clients that support remote MCP OAuth. See llms-install.md for step-by-step client setup and the MCP guide for the complete tool list.

Puppetflow in action

🛟 Discussion / Need help?

Join our Discord

Open an Issue

Project philosophy

Puppetflow is built around a few straightforward ideas:

  • Browser automation should remain understandable and debuggable.
  • Self-hosting should be a real option, not an afterthought.
  • Good defaults matter, but users should keep control of their data and infrastructure.
  • Reliability and maintainability matter more than shipping a long list of half-finished features.

A note on project history

The public Git history was intentionally reset for the Community Edition release for security reasons. This repository's commit count therefore does not reflect the age or maturity of the project.

Puppetflow has been developed and operated in real conditions for nearly a year. The Community Edition is a publication of an existing product, not the beginning of a prototype.

No AI Slop Policy

Puppetflow does not accept low-effort, unreviewed generated code or content.

AI-assisted work is not banned. Submitting output that the contributor does not understand, has not checked, or cannot maintain is. Contributions must be deliberate, consistent with the existing codebase, and owned by the person submitting them.

See Contributing for the practical rules.

Repository documentation

License

Puppetflow is source available. Most of the code is licensed under the Puppetflow Source Available License, which permits personal, non-commercial use. Internal business use requires a current paid Puppetflow subscription or order.

Files under proprietary/, files containing the .pp. marker, and explicitly marked code blocks implement paid features. They are covered by the Puppetflow Proprietary License, and their use in production requires a valid Puppetflow subscription.

Frequently asked questions

What is Puppetflow?

Build, run, monitor, and control Puppetflow browser automations from AI clients.

How do I install Puppetflow?

It runs as a remote server. In Claude Code: claude mcp add --transport http puppetflow 'https://{instance_host}/api/workspaces/{workspace}/mcp-server/http'. For other clients, paste the JSON configuration from this page into their MCP settings.

Does Puppetflow need an API key?

Its registry entry lists no required key or secret. Some features may still need an account with the service it connects to.

Is Puppetflow open source?

Its code is public on GitHub, but the repository does not declare a standard license.

Who maintains Puppetflow?

It is published by puppetflow. AI.tn lists it but does not maintain or endorse it.

From the official MCP Registry. Listing is not an endorsement: review a server's code and permissions before giving it access to your data.