trackmcp
Back to directory
DevinoSolutions

stealth-chrome-devtools-mcp

View on GitHub

Undetectable browser automation for AI agents via MCP. Stealth Chrome DevTools with smart profile management, anti-detection arg filtering, and CDP access.

26 stars PythonOthers Updated Sep 2, 2026
anti-detectionbrowser-automationcdpchrome-devtoolsmcpmodel-context-protocolnodriverpythonstealthweb-scraping

Documentation

Stealth Chrome DevTools MCP

PyPI
Tests
Python 3.11+
License: AGPL-3.0
MCP

> Undetectable browser automation for AI agents via the Model Context Protocol.

A self-contained stealth Chrome DevTools MCP server with smart profile management, anti-detection stealth arg filtering, and robust process lifecycle handling. Built on nodriver (CDP-based) for full anti-bot evasion.


Demos

Cloudflare Turnstile Bypass

https://github.com/user-attachments/assets/c4de61ae-6878-4fff-9bfd-65cdd4fadc2f

Watch on YouTube

Persistent Login Sessions

https://github.com/user-attachments/assets/f81fc0c2-9233-48cd-8a9d-2577b1d33d57

Watch on YouTube


Key Features

  • Undetectable by anti-bot systems — Cloudflare, DataDome, PerimeterX, etc.
  • Smart profile management — master/snapshot/clone strategy preserves logins across sessions
  • Stealth arg filtering — automatically strips 30+ detectable Chrome flags (Puppeteer/Playwright signatures, automation markers)
  • Multi-instance support — spawn and manage multiple browsers simultaneously
  • A shared backend across sessions — every client session proxies to a shared

backend process rather than starting its own, one per desktop so a headed spawn

lands on a real screen; simultaneous cold start is scale-tested at 40 concurrent

sessions, all usable in seconds against one backend

  • Auto-suffix busy profiles — `github-session` auto-becomes `github-session-2` when occupied
  • Orphan recovery — safely cleans up leaked browser processes without killing live ones
  • Session persistence — cloned profiles carry cookies, logins, and Web Data from master
  • Zero idle timeout — browsers stay alive until explicitly closed
  • Full CDP access — DOM manipulation, network interception, JavaScript execution, screenshots

Installation

The right way — `uv tool install` (persistent, fleet-safe)

bash
uv tool install stealth-chrome-devtools-mcp==2.1.1

This installs a version-pinned executable at `~/.local/bin/stealth-chrome-devtools-mcp`

(Windows: `%USERPROFILE%\.local\bin\stealth-chrome-devtools-mcp.exe`). Point your MCP

config (`claude_desktop_config.json`, `~/.claude.json`, etc.) at it:

json
{
  "mcpServers": {
    "stealth-chrome-devtools-mcp": {
      "command": "C:\\Users\\\\.local\\bin\\stealth-chrome-devtools-mcp.exe",
      "args": []
    }
  }
}

Claude Code one-liner (use the `.exe` path above on Windows):

bash
claude mcp add --scope user stealth-chrome-devtools-mcp -- ~/.local/bin/stealth-chrome-devtools-mcp

Why not `uvx` in the config? It works, but `uvx` re-resolves the package on

every client session start. Each Claude Code session launches its own stdio

proxy, so a fleet of concurrent sessions (the shared backend is scale-tested at

40) turns startup into a package-resolution storm. A `uv tool install` gives

every proxy an instant, pinned executable — the shared backend, profile

handling, and per-session browser isolation behave identically.

To upgrade later: `uv tool install stealth-chrome-devtools-mcp==`

(or `uv tool upgrade stealth-chrome-devtools-mcp` to track the latest release).

Alternatives

Zero-install trial (fine for a first look, not for fleets):

json
{
  "mcpServers": {
    "stealth-chrome-devtools-mcp": {
      "command": "uvx",
      "args": ["stealth-chrome-devtools-mcp==2.1.1"]
    }
  }
}

Or via pip (`pip install stealth-chrome-devtools-mcp==2.1.1`), then use the

`stealth-chrome-devtools-mcp` console script from that environment as the

`command`.

Crashes are reported to the maintainers by default, with your username and

machine name scrubbed out. See Error Reporting for what a

report contains and how to turn it off.

Local Development

json
{
  "mcpServers": {
    "stealth-chrome-devtools-mcp": {
      "command": "uv",
      "args": [
        "--directory", "/path/to/stealth-chrome-devtools-mcp",
        "run", "stealth-chrome-devtools-mcp"
      ]
    }
  }
}

How It Works

Browser Profile Strategy

code
C:\stealth-mcp-browser-sessions\
  master/              # Your primary Chrome profile (logins, cookies, extensions)
  master-snapshot/     # Safe copy refreshed while master is closed
  sessions/            # Cloned profiles for concurrent use
    github-session/
    github-session-2/  # Auto-suffixed when github-session is busy

1. `spawn_browser()` uses the master profile when available

2. Before opening master, the server refreshes `master-snapshot`

3. When master is busy, a clone is created from the snapshot

4. Clones carry all cookies, logins, and session data

5. Stale snapshots are auto-refreshed when auth files change

Clones exclude regenerable Chrome caches, so each is a few MB rather than

multiple GB. Disposable auto-clones are deleted on close, and a storage cap

(`STEALTH_MCP_CLONE_STORAGE_CAP_GB`, default 10 GB) reclaims the oldest idle

clones if any ever leak — so `sessions/` stays bounded. Cap eviction is

recoverable: an evicted clone is moved into `sessions/.trash/` and only

purged after a retention window (`STEALTH_MCP_CLONE_TRASH_RETENTION_HOURS`,

default 24 h), so a mistaken eviction can be restored rather than lost.

Named profiles you create explicitly (e.g. `github-session`) persist and are

never deleted. But even a "persistent" profile is ~98% regenerable (caches plus

Chrome's multi-GB on-device AI model). So when `sessions/` exceeds

`STEALTH_MCP_BROWSER_SESSION_STORAGE_CAP_GB` (default 20 GB), the largest idle named

profiles are trimmed of those regenerable dirs while **every login is

preserved** — Chrome rebuilds them on next launch. In-use profiles are never

touched.

> Shared-machine note: the browser-session root defaults to `C:\stealth-mcp-browser-sessions`

> (drive root), which holds your logged-in cookies and session data. On a

> single-user machine this is fine. On a shared multi-user Windows box, other

> local users may be able to read it — point `STEALTH_MCP_BROWSER_SESSION_ROOT`

> at a location inside your user profile (e.g. `%LOCALAPPDATA%\stealth-mcp`) so

> the OS user ACLs protect it.

Stealth Arg Filtering

The server automatically strips Chrome flags that would compromise stealth:

CategoryExamplesWhy Stripped
Automation signals`--enable-automation`, `--test-type`Sets `navigator.webdriver=true`
Fingerprint leaks`--disable-gpu`, `--disable-webgl`Detectable via WebGL/canvas probes
Puppeteer defaults`--disable-backgrounding-occluded-windows`Bot signature fingerprint
Playwright defaults`--password-store=basic`, `--use-mock-keychain`Bot signature fingerprint

Stripped args are reported in `spawn_diagnostics.stealth_args_stripped`.

Orphan Recovery

On server restart, the process cleanup system:

  • Reaps only browsers whose owner backend is dead — every tracked browser records

which backend started it, so two backends running side by side never reap each

other's browsers

  • Keeps `create_time` tracking as a second net: never kills a process that started

after the current server session began

  • Safely handles `psutil.AccessDenied` on Windows elevated processes

Headed Browsing and Where the Window Opens

A headed browser appears on the desktop of whichever process launched it, not

of whichever session asked. Because sessions share a backend, a backend that was

first started from an SSH login or a Windows service session cannot show a window

to anyone — including the sessions running on the physical desktop.

So the backend is keyed by display context: one per desktop, plus one for a

headless context. Discovery prefers a backend that can show a window, which means

an SSH-driven `spawn_browser(headless=False)` automatically uses the desktop

backend and its window opens on the real screen. Where no such backend exists, the

spawn raises instead of handing back an invisible browser; run

`stealth-chrome-devtools doctor` to see which contexts have a backend. Headless

spawns work from anywhere.

Usage Examples

python
# Spawn with default master profile
spawn_browser()

# Named session with login persistence
spawn_browser(user_data_dir="github-session")

# Same name while first is open → auto-suffixes to github-session-2
spawn_browser(user_data_dir="github-session")

# Headless with stealth (bad args auto-stripped)
spawn_browser(headless=True, browser_args=["--enable-automation"])
# → stealth_args_stripped: ["--enable-automation stripped: sets navigator.webdriver=true"]

MCP Tools

ToolDescription
`spawn_browser`Launch a new stealth browser instance
`navigate`Navigate to a URL
`take_screenshot`Capture page screenshot
`execute_script`Run JavaScript in page context
`query_elements`Find DOM elements by CSS selector
`click_element`Click on an element
`type_text`Type text into an input
`get_page_content`Get page HTML content
`list_instances`List all active browser instances
`close_instance`Close a specific browser
`list_network_requests`View intercepted network traffic
`get_cookies` / `set_cookie`Manage browser cookies

94 tools across 11 sections — the count is derived from the live tool registry,

never hand-maintained. See the full navigation map →.

That is what the server serves, which is not the same as what the release

gate proves. At the release SHA in the evidence ledger, 3 of those 94 are

release-qualified: asserted end-to-end over the real stdio transport a client

actually speaks. The rest are driven against real Chrome by the E2E suite but

through an in-process seam, so they are `served-unqualified` at the wire — tested,

not proved there. `RELEASE_CONTRACT.md` lists the state of

each tool and is the only source for those numbers.

Testing

bash
# Unit tests only (no Chrome needed)
uv run pytest -m "not integration"

# All tests (needs Chrome installed)
uv run pytest

# Verbose with short tracebacks
uv run pytest -v --tb=short

> If your checkout path contains spaces or an `&`, `uv run pytest` fails with

> `Failed to canonicalize script path` — use the venv Python directly:

> `.venv\Scripts\python.exe -m pytest -m "not integration"`. See

> CONTRIBUTING.md for the full test/gate workflow.

A comprehensive suite covers stealth arg filtering, profile resolution, orphan recovery, storage-cap sweeps, the ops CLI, and full browser integration.

Environment Variables

All optional. Defaults work for normal use. Set them in your shell, or in

`~/.stealth-mcp/.env` — every key is documented in `.env.example`.

A `.env` in your project directory is deliberately ignored. The backend is a

shared process launched with whatever folder your MCP client had open, so

reading the project's `.env` meant reading someone else's application config —

which crashed the server outright on an ordinary `DATABASE_URL` and silently

adopted that app's `PORT`, `DEBUG`, and `SENTRY_DSN` as the server's own.

VariableDefaultPurpose
`STEALTH_MCP_BROWSER_SESSION_ROOT``C:\stealth-mcp-browser-sessions` (Win) / `~/.stealth-mcp-browser-sessions` (Unix)Base folder for profiles
`BROWSER_MASTER_USER_DATA_DIR``/master`Master Chrome profile path
`BROWSER_MASTER_SNAPSHOT_DIR``/master-snapshot`Snapshot clone source
`BROWSER_PROFILE_CLONE_ROOT``/sessions`Folder for profile copies
`BROWSER_PROFILE_REFRESH_DAYS``7`Refresh copies after N days (`0` = disable)
`STEALTH_MCP_CLONE_STORAGE_CAP_GB``10`Cap on total auto-clone storage; oldest idle clones are reclaimed when exceeded (`0` = disable). Named profiles and in-use clones are never touched.
`STEALTH_MCP_BROWSER_SESSION_STORAGE_CAP_GB``20`Cap on total `sessions/` storage; when exceeded, the largest idle named profiles are trimmed of regenerable cache/model dirs — logins kept (`0` = disable). *(Renamed from `STEALTH_MCP_SESSION_STORAGE_CAP_GB`; update your config — the old name is no longer read.)*
`STEALTH_MCP_CLONE_TRASH_RETENTION_HOURS``24`How long a cap-evicted clone stays recoverable in `sessions/.trash/` before purge (`0` = purge on next sweep).
`STEALTH_MCP_CLONE_OUTPUT_DIR``~/.stealth-mcp/element_clones`Where screenshots, large-response spills, and element-clone files are written. Kept in a per-user dir (never inside the installed package) so a read-only `site-packages` can't break captures.
`BROWSER_IDLE_TIMEOUT``0`Idle cleanup timeout (`0` = disabled)
`STEALTH_CHROME_PROFILE_KEY`unsetForce a stable clone key
`STEALTH_MCP_CLIENT_ROOTS_TIMEOUT_SECONDS``5`Deadline for the `roots/list` request the auto-clone path sends to the MCP client to name a clone. MCP `roots` is optional, so a client may never answer; on expiry the clone name falls back to `CODEX_WORKSPACE`/`CLAUDE_PROJECT_DIR`/`PWD`/cwd (`0` = never ask).
`STEALTH_BROWSER_DEBUG``false`Enable debug logging
`STEALTH_MCP_NO_ERROR_REPORTING``false`Set to `true` to disable error reporting

CLI

Installs a `stealth-chrome-devtools` ops command for managing the server and its

disk usage. (This is for *ops* — to drive a browser, use the MCP server or its

HTTP backend.)

These four only read and preview — they change nothing, and the test suite runs

them on every commit, so they are known to work:

console
stealth-chrome-devtools status
stealth-chrome-devtools profiles
stealth-chrome-devtools cleanup
stealth-chrome-devtools cleanup --browser-session-cap-gb 12

`status` reports whether the backend is up plus the browser-session root and both

caps; `profiles` lists profiles with size / role / in-use; `cleanup` previews the

reclaimable disk (dry run), and `--browser-session-cap-gb` previews it at a

tighter cap.

These are not auto-executed — `--apply` deletes, `serve` does not return, and

`doctor` needs Chrome installed:

console
stealth-chrome-devtools cleanup --apply               # actually reclaim
stealth-chrome-devtools doctor                        # check Chrome / environment
stealth-chrome-devtools serve --http --port 19222     # start the server

`cleanup` deletes idle auto-clones over the clone cap and trims idle named

profiles down to their session state — logins kept — over the browser-session cap. It

is a dry run unless you pass `--apply`, never touches in-use profiles, and

uses the same selectors as the automatic sweep, so the preview matches `--apply`.

Preparing the Master Profile

1. Start the MCP server

2. Call `spawn_browser()` without `user_data_dir`

3. Sign in to your accounts in the browser that opens

4. Close it — future sessions use this profile or clone from it

Requirements

  • Python 3.11+
  • Chrome, Chromium, or Microsoft Edge
  • uv (recommended) or pip
  • A desktop session for headed browsing (headless works from SSH, CI, and services)

Error Reporting

Crashes and errors are reported to Sentry by default, so

that a failure you hit is a failure we can see and fix. There is nothing to

install and nothing to configure: the SDK ships with the package and the

destination is built in.

What a report contains. The exception type and message, the stack trace, the

package version, and the platform. Three things are kept out of it:

  • your machine name (Sentry's `server_name`) is dropped entirely;
  • your username is removed from every path, so a stack frame reads

`C:\Users\~\...`, `/home/~/...` or `/Users/~/...` instead of your home

directory;

  • local variables are not captured at all. The Sentry SDK sends them by

default; we turn that off, because a local in this tool can hold a proxy

password, an `Authorization` or `Cookie` header, or a script you asked it to

run — secrets that no path rule could rescue.

That is universal — it runs on every install, ours included, and there is no way

to opt back into sending those fields. What it deliberately leaves alone is the

part that makes a report useful: the error type, the module path after the home

segment, the source line that failed, and the release it came from.

An error message still quotes whatever the failing call was working with — a URL

you navigated to, a file you asked for. If that is not a trade you want to make,

turn reporting off.

To turn it off, set one variable in your shell or in `~/.stealth-mcp/.env`:

bash
STEALTH_MCP_NO_ERROR_REPORTING=true

Earlier releases read `SENTRY_DSN` from the environment. They no longer do —

that variable belongs to *your* application, and a shared backend launched from

your project folder was picking it up. See

Environment Variables for why this tool ignores your

project's `.env` entirely.

Development setup

bash
git clone https://github.com/DevinoSolutions/stealth-chrome-devtools-mcp
cd stealth-chrome-devtools-mcp
uv sync --extra dev --extra test   # install linters + test deps
npm install                        # arm husky pre-commit/pre-push hooks

The six quality gates run automatically on every commit:

ruff format, ruff check, ty check, vulture, suppression-owner check, file-budget check.

Unit tests run on pre-push.

Documentation

  • **CLAUDE.md** — navigation map of the source tree + glossary + conventions
  • **DESIGN.md** — architecture invariants and the *why* behind them
  • **RUNBOOK.md** — operating the backend: verbs, logs, recovery, MCP smoke path
  • **CONTRIBUTING.md** — clone → install → test, the quality gate, conventions

License

See LICENSE.


Built by Devino Solutions

Frequently asked questions

What is stealth-chrome-devtools-mcp?

stealth-chrome-devtools-mcp is Undetectable browser automation for AI agents via MCP. Stealth Chrome DevTools with smart profile management, anti-detection arg filtering, and CDP access.

How do I install stealth-chrome-devtools-mcp?

Open the GitHub repository and follow its README. Most MCP servers are added to your client's MCP config, then called by your agent.

Is stealth-chrome-devtools-mcp open source?

Yes — it is hosted on GitHub at https://github.com/DevinoSolutions/stealth-chrome-devtools-mcp and has 26 stars.

Related MCP tools

Run your own MCP server? See who uses it and what to fix.

Measure it with TrackMCP