LABSCO SUMMARY

The author's warning against public deployment is the most useful line in the README.

A toolkit that drives a browser, scrapes pages and reads developer-tools output is powerful locally and dangerous exposed, and most projects with this surface never say so. Saying it plainly means the boundary is a design decision rather than something you discover. Treat it as a workstation tool: it belongs next to the browser it drives, not on a server.

What it is

A Python toolkit covering browser automation, AI analysis, visual processing, web scraping, automated test generation and developer-tools analysis. Its own documentation states it is designed for local single-user use and is not suitable for public deployment.

What you get

Driving a browser and capturing what it shows
Visual processing and AI analysis over the result
Web scraping
Generating automated tests
Reading developer-tools data from the page

Requirements

Python 3.12 or newer and uv, with API keys supplied through a local environment file. MIT licensed. Documentation is in Chinese.

Setup effort

Build from source — clone the repository, then follow the steps on this page and add the configuration for your client

See installation options

USE CASES

01

Driving a page and reading it back

Automation plus visual and developer-tools analysis together.

02

Generating tests from behaviour

Automated test generation over what the browser did.

6 tools

Listed in project docs. Your client may show a different list.

open_browser

click_element

fill_form

take_screenshot

assert_element_present

get_console_logs

BEFORE YOU INSTALL

What to know first

Tools you get

6 tools named in the project's documentation.

Requires

uv on PATH
Python 3.12+
OPENAI_API_KEY in your environment
ANTHROPIC_API_KEY in your environment
GOOGLE_API_KEY in your environment

Where it runs

Both documented clients run the server on your machine.

INSTALLATION

Connect your client

Clone the repository, then follow the steps on this page and add the configuration for your client. This one is built from source rather than installed with a single command.

Server setup

Setup

TERMINAL

git clone https://github.com/0x-hewm/percepta-mcp
cd percepta-mcp
uv venv
uv sync
cp .env.example .env
uv run playwright install chromium

uv

TERMINAL

uv run python -m src.percepta_mcp.server

MCP Clients

Claude Desktop

CLAUDE DESKTOP · JSON

{
  "mcpServers": {
    "percepta-mcp": {
      "command": "uv",
      "args": [
        "run",
        "python",
        "-m",
        "src.percepta_mcp.server"
      ],
      "cwd": "<YOUR_PATH_TO_PERCEPTA_MCP>"
    }
  }
}

The configuration on this page does not set OPENAI_API_KEY, ANTHROPIC_API_KEY and GOOGLE_API_KEY; the project's instructions put them in a .env file. Check which folder the server reads that file from before you rely on it.

VS Code

VS CODE · JSON

{
  "servers": {
    "percepta-mcp": {
      "command": "uv",
      "args": [
        "run",
        "python",
        "-m",
        "src.percepta_mcp.server"
      ],
      "cwd": "${workspaceFolder}"
    }
  }
}