SkillFoxx

Electron Debug MCP

Electron debugging MCP server: 111 tools let an agent see the window, click buttons, read console and network and audit security.

  • 102 ★ on the repository
  • Last commit
  • ISC
MCP server

Extended access

Why this level

  • The server reads and changes the state of a running app: DOM, storage, clipboard and permissions.
  • For debugging the app starts with an open local port.
All reasons and checks

amafjarkasi/electron-mcp-serverAuthor:amafjarkasi

Install

In your terminal, with SkillFoxx CLI

npx skillfoxx add mcp/electron-debug-mcp

Detects the agents on your machine, shows the access level and pins the version.

Other ways to install

Access and checks

How we review

Extended accessThe tool runs code or makes network calls and only reads external data.

  • The server reads and changes the state of a running app: DOM, storage, clipboard and permissions.
  • For debugging the app starts with an open local port.

This is third-party code. Review the repository files before installing.

What it does

The server connects an agent to a running Electron app through the Chrome DevTools Protocol and the Node inspector. The agent launches the app with a debug port or attaches to a running one, then gets a screenshot, the accessibility tree, console errors and network failures, and clicks and types. Main-process tools show IPC calls and infer the exchange schema as TypeScript interfaces, mock native Electron modules such as `dialog` or `shell`, and hunt memory leaks. Composite flows `act_and_verify` and `triage` cut round trips. The `electron_security_audit` tool checks windows against Electron guidance: `nodeIntegration`, context isolation, sandbox and CSP.

Who it is for. Developers of Electron desktop apps who want Claude Code or Cursor to run and verify their own work in the window.

Good fit when

  • An agent changed the code and the result must be seen in the window by clicking buttons and reading the console.
  • The main process crashes silently and the exception trace is needed.
  • Window security settings are checked against Electron guidance before a release.
  • The IPC exchange schema between the window and the main process must be captured.

Not a fit when

  • The app runs in a browser rather than Electron; Chrome DevTools MCP fits better.
  • A debug port cannot be opened on the work machine.
  • Two or three of the 111 tools are enough and the set would overload agent context.

Example request

Launch my Electron app, open the settings screen, press Save and show console errors and network failures. Then run `electron_security_audit`.

MCP

Transport
stdio
Authentication
not required

Limitations

The server sees and changes everything reachable through the debug port: DOM, storage, clipboard, permissions and IPC handlers. An open debug port is reachable by other local processes, so do not leave it open permanently. For the main process the app must be launched with `inspectMain: true`. Secrets are substituted as `{{secret:VAR}}` from the environment and masked in output, but a password typed into a form still reaches the app. Node.js 20 or newer is required and the license is ISC.

How to disable. Remove the `electron-debug` entry from the MCP client configuration and close any app launched with a debug port.

FAQ

How do I connect it to Cursor?

Add an `electron-debug` server to `.cursor/mcp.json` with command `npx` and arguments `-y`, `electron-debug-mcp@latest`. For Claude Desktop the same block goes into `claude_desktop_config.json`.

How do I debug the main process?

Launch the app through `start_app` with `inspectMain: true`, which opens the Node inspector port. Without it the main-process tools return an error.

How do I type a password without leaking it into logs?

Pass `{{secret:NAME}}` instead of the text: the value comes from `process.env` and is replaced by `[REDACTED]` in output, console logs and HAR files.

Can it attach to an app that is already running?

Yes: `attach` by DevTools port or WebSocket URL, and `discover_apps` scans local ports 9222 through 9322.

Can a scenario be recorded for Playwright?

Yes, the composite `run_script` tool exports the executed steps in Playwright format.

Discussion

No questions or reviews yet.

MCP servers
Official

An MCP server from Mozilla: an agent drives Firefox through WebDriver BiDi, reads the console and network, takes page snapshots and profiles loading

MCP serverBroad access567
Editors’ pick

The official MCP server debugger: web UI, automation CLI and terminal UI in one package

CLIExtended access11.3K
Official

An agent gets a live session of your web, iOS or Android app, sees the screen, writes tests and checks its own edits.

MCP serverExtended access13.8K

An MCP server that checks whether the agent is really done: drives the real app and returns pass, fail or couldn't tell with the file and line

MCP serverExtended access1.3K
Official

kogiQA MCP gives an agent a browser that finds page elements without selectors, for debugging web apps and writing e2e tests

MCP serverExtended access88

CLI, MCP server and typed Node.js API so an agent can verify its changes in a live mobile app on iOS, Android and more

CLIExtended access5.4K

README in short

Tools are grouped by connection, observation, actions, network, storage, main process, security and diagnostics. There are 9 resources and 6 ready prompts, and `doctor` checks the environment.

README badge

Report a problem