Skip to main content
Glama

Debugg AI MCP

Official
by debugg-ai

🧪 Official MCP Server for Debugg AI

AI-driven browser automation and E2E test server implementing the Model Context Protocol (MCP), designed to help AI agents test UI changes, simulate user behavior, and analyze visual outputs of running web applications — all via natural language and CLI tools.

End to end testing used to be a nightmare. Not just to setup, but to manage over time as you made changes to your app.

Debugg AI's MCP server offers a NEW way to test, where you never have to worry about setting up playwright, local browsers or proxies with it fully remote, managed browsers that simply connect to a server running locally or remotely via a secure tunnel.

That means no distracting chrome pop ups as it's running tests, no managing chrome or playwright versions, and best of all - ZERO CONFIGURATION. Just grab an API key and add us to your MCP server list.

Should you want to later rerun those tests or create a suite of them to run in your CI / CD pipeline, you can see all historical test results in your dashboard - Debugg.AI App


🚀 Features

  • 🧠 MCP Protocol Support Full MCP server implementation with CLI and tool registry support.
  • 🧪 End-to-End Test Automation Trigger UI tests based on user stories or natural language descriptions via the debugg_ai_test_page_changes tool.
  • 🌐 Localhost Web App Integration Test your running dev app on any localhost port with simulated user flows.
  • 🧾 MCP Tool Notifications Sends real-time progress updates back to clients with step descriptions and UI state goals.
  • 🧷 Screenshot Support Capture final visual state of the page for LLMs with image rendering support.
  • 🧱 Stdio Server Compatible Plug into any MCP-compatible client (like Claude Desktop, LangChain agents, etc.) via stdin/stdout.

Examples

Input prompt: "Test the ability to create an account and login"

Test Create Account and Login

Results:

**Task Completed** - Duration: 86.80 seconds - Final Result: Successfully completed the task of signing up and logging into the account with the email 'alice.wonderland1234@example.com'. - Status: Success

Full Demo:

Watch a more in-depth, Full Use Case Demo


🛠️ Quickstart

Ensure you have created a free account and generated an API Key - DebuggAI

Option 1: NPX (Local Development)

npx -y @debugg-ai/debugg-ai-mcp

Use this when testing or integrating into tools like Claude Desktop or your own AI agent.

Option 2: Docker

docker run -i --rm --init \ -e DEBUGGAI_API_KEY=your_api_key \ -e TEST_USERNAME_EMAIL=your_test_email \ -e TEST_USER_PASSWORD=your_password \ -e DEBUGGAI_LOCAL_PORT=3000 \ -e DEBUGGAI_LOCAL_REPO_NAME=your-org/your-repo \ -e DEBUGGAI_LOCAL_BRANCH_NAME=main \ -e DEBUGGAI_LOCAL_REPO_PATH=/app \ -e DEBUGGAI_LOCAL_FILE_PATH=/app/index.ts \ quinnosha/debugg-ai-mcp

🧰 MCP Tool: debugg_ai_test_page_changes

Description

Run an end-to-end test on a running web app, testing a UI feature or flow described in natural language. Allows AI agents in ANY code gen platform to quickly evaluate proposed changes and ensure new functionality works as expected.

Input Parameters

NameTypeRequiredDescription
descriptionstringWhat feature or page to test (e.g. "Signup page form")
localPortnumberPort of your running app (default: 3000)
repoNamestringGitHub repo name
branchNamestringCurrent branch
repoPathstringAbsolute path to the repo
filePathstringFile to test

🧪 Example Claude Desktop Config

{ "mcpServers": { "debugg-ai-mcp": { "command": "npx", "args": ["-y", "@debugg-ai/debugg-ai-mcp"], "env": { "DEBUGGAI_API_KEY": "YOUR_API_KEY", "TEST_USERNAME_EMAIL": "test@example.com", "TEST_USER_PASSWORD": "supersecure", "DEBUGGAI_LOCAL_PORT": 3000, "DEBUGGAI_LOCAL_REPO_NAME": "org/project", "DEBUGGAI_LOCAL_BRANCH_NAME": "main", "DEBUGGAI_LOCAL_REPO_PATH": "/Users/you/project", "DEBUGGAI_LOCAL_FILE_PATH": "/Users/you/project/index.ts" } } } }

🔐 Environment Variables

VariableDescriptionRequired
DEBUGGAI_API_KEYAPI key for calling DebuggAI backend
TEST_USERNAME_EMAILEmail of test user account
TEST_USER_PASSWORDPassword of test user account
DEBUGGAI_LOCAL_PORTLocal port your app runs on
DEBUGGAI_LOCAL_REPO_NAMEGitHub repo name
DEBUGGAI_LOCAL_BRANCH_NAMEBranch name
DEBUGGAI_LOCAL_REPO_PATHLocal path to repo root
DEBUGGAI_LOCAL_FILE_PATHFile to test

🧑‍💻 Local Development

# Clone the repo and install dependencies npm install # Copy the test config and insert your creds cp test-config-example.json test-config.json # Run the MCP server locally npx @modelcontextprotocol/inspector --config debugg-ai-mcp/test-config.json --server debugg-ai-mcp

📁 Repo Structure

. ├── e2e-agents/ # E2E browser test runners ├── services/ # Client for DebuggAI API ├── tunnels / # Secure connections to remote web browsers ├── index.ts # Main MCP server entry ├── Dockerfile # Docker build config └── README.md

🧱 Built With


💬 Feedback & Issues

For bugs, ideas, or integration help, open an issue or contact the DebuggAI team directly.


🔒 License

MIT License © 2025 DebuggAI


Related MCP Servers

View all related MCP servers

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/debugg-ai/debugg-ai-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server