Openbrowser AI runs on your own machine — the client starts it, so there is no endpoint to ping. 324 installs a week from pypi. Last commit 28 Sep 2026.
AI browser automation. Write async Python to navigate, click, type, and extract data.
We read the source, 5 d ago · rules 995398eba69f
Things with no honest explanation: a promise that contradicts the code, code that runs at install time while hiding what it does, data leaving the machine.
curl -LsSf https://astral.sh/uv/install.sh | sh
TOKEN=$(curl -s -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-token-ttl-seconds: 60")
TOKEN=$(curl -s -X PUT "http://169.254.169.254/latest/api/token" -H "X-aws-ec2-metadata-token-ttl-seconds: 60")
What this server is able to do. For an MCP server this is often the job itself — a terminal server runs commands because that is what it is for. Listed so you know what you are plugging in, not as an accusation.
if Path('/.dockerenv').exists() or 'docker' in Path('/proc/1/cgroup').read_text().lower():
proc = subprocess.Popen(command, stdin=subprocess.PIPE, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
download_path = os.path.join(downloads_dir, final_filename)
exec(compiled, self._namespace)
RE_SELECT = re.compile(_P + r"Select ['\"](.+?)['\"] from the ['\"](.+?)['\"] field")
dropdown.innerHTML = results.map(item => options.template(item)).join('');
Is this your server and something here is wrong? Tell us — corrections are free and do not require a plan.
Code changes quietly between releases, and nobody reads the diff of a dependency. We do, on every release — watch Openbrowser AI and you get told the day something new turns up.
This server runs on your own machine — install it with the package manager and the client starts it for you. Package name taken from the official registry entry.
claude mcp add openbrowser-ai -- uvx openbrowser-ai
{
"mcpServers": {
"openbrowser-ai": {
"command": "uvx",
"args": [
"openbrowser-ai"
]
}
}
}
[mcp_servers.openbrowser-ai]
command = "uvx"
args = ["openbrowser-ai"]
{
"mcpServers": {
"openbrowser-ai": {
"command": "uvx",
"args": [
"openbrowser-ai"
]
}
}
}
{
"servers": {
"openbrowser-ai": {
"type": "stdio",
"command": "uvx",
"args": [
"openbrowser-ai"
]
}
}
}
Optional settings from the registry entry, left out of the block above:
OPENBROWSER_HEADLESS (Run browser in headless mode (true/false)), OPENBROWSER_ALLOWED_DOMAINS (Comma-separated list of allowed domains (empty for no restrictions)), OPENBROWSER_USER_DATA_DIR (Optional explicit Chrome profile directory. Omit for an isolated profile per MCP server instance; concurrent owners of one explicit profile are rejected), OPENBROWSER_STORAGE_STATE (Optional JSON file used as a one-time seed for cookies plus localStorage in managed MCP profiles; later updates are saved to the active profile copy), OPENBROWSER_COMPACT_DESCRIPTION (Use minimal tool description (~500 tokens instead of ~1,430)), OPENBROWSER_MAX_OUTPUT (Maximum output characters per code execution (default: 10000)), ANONYMIZED_TELEMETRY (Enable anonymized usage telemetry (true/false)).
AI browser automation - navigate, click, type, extract content, and run autonomous web tasks
AI-powered browser automation — navigate, click, fill forms, and extract data from any website.
Give AI agents full control over live browsers and apps to navigate, click, extract, and automate.
Automate a browser with accessibility snapshots: navigate, click, type, fill and wait.
Browser automation: navigate, click, fill, screenshot and evaluate in a headless browser.
Browser automation MCP server: navigate, fill forms, click, screenshot, and manage tabs
AI-powered browser automation MCP — 6 tools for web interaction, extraction, and validation.
Lightweight browser automation for AI agents: navigate, screenshot, extract, PDF. One small binary.
Answers built from our own checks of this server.