cliqueja
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@cliquejaclick the 'Accept cookies' button and snapshot the page"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Clique Já
Why it exists
Playwright and puppeteer are great — and heavy for “just click that button now”.
Clique Já grabs a Chrome tab you already use, injects a red micro-runtime (__cliqueja) and responds to commands via WebSocket in milliseconds.
Clique Já | Classic automation | |
Browser | Real Chrome (your session) | Isolated Chromium |
Session / login | already logged in | re-auth / storage state |
Latency per click | 1 | full stack |
AI agent | MCP tools | scripts / custom |
Related MCP server: monkeysee
Flowchart
flowchart LR
subgraph Agent["Agente / CLI"]
MCP["mcp.mjs<br/>tools MCP"]
HTTP["POST /cmd"]
end
subgraph Hub["Hub local :17321"]
H["hub.mjs<br/>WebSocket + HTTP"]
end
subgraph Chrome["Chrome"]
EXT["Extensão Clique Já<br/>service worker"]
TAB["Aba http(s)"]
INJ["inject.js<br/>window.__cliqueja"]
end
MCP -->|WS| H
HTTP -->|HTTP| H
H <-->|WS| EXT
EXT -->|scripting MAIN world| INJ
INJ --> TABFlow of a click
sequenceDiagram
participant A as Agente
participant H as Hub :17321
participant E as Extensão
participant P as Página
A->>H: { method: "click", sel: "button" }
H->>E: mesmo envelope (WS)
E->>P: inject + __cliqueja.click
P-->>E: ok (+ flash vermelho)
E-->>H: { ok: true }
H-->>A: resultDetailed architecture: docs/ARCHITECTURE.md
Quick start
1. Hub
git clone https://github.com/paivaxqz/cliqueja.git
cd cliqueja
npm install
npm run hub
# ou: start-hub.bat (Windows)Hub listens only on 127.0.0.1:17321.
2. Extension
Chrome →
chrome://extensions→ Developer modeLoad unpacked → folder
ext/Open an
http(s)site (notchrome://)Click the icon → use this tab (red ON badge)
3. Smoke HTTP
curl http://127.0.0.1:17321/status
curl -X POST http://127.0.0.1:17321/cmd -H "content-type: application/json" -d "{\"method\":\"ping\"}"4. MCP in AI (user's PC)
Hub + extension run on the user's machine — not in the cloud.
Full guide (3 steps + Cursor/Claude): docs/MCP.md
Cursor config — path of the user's clone:
{
"mcpServers": {
"cliqueja": {
"command": "node",
"args": ["${workspaceFolder}/mcp.mjs"]
}
}
}If the folder opened in Cursor is not the clone, use an absolute path (C:\\Users\\...\\cliqueja\\mcp.mjs or /Users/.../cliqueja/mcp.mjs).
Tools: cliqueja_status · abas · usar_aba · ir · click · fill · texto · snapshot · links · upload
Why not Vercel?
Vercel doesn't open the user's Chrome nor receive the extension's WebSocket (127.0.0.1).
Landing/docs on Vercel ok. Hub + MCP + extension = always local. → docs/MCP.md
5. Lib mode (thin Playwright)
npm run exemploindex.js + inject.js: Chrome via playwright-core, 1 evaluate per action, red mark in the corner.
Selectors
CSS:
button.buy,#emailText:
text=EntrarXPath:
xpath=//button[@type='submit']
Extension default timeout: 4s (inject waits with requestAnimationFrame).
Structure
cliqueja/
├── ext/ # Chrome MV3 (load unpacked)
│ ├── manifest.json
│ ├── background.js # WS client + comandos
│ ├── inject.js # runtime na página
│ ├── popup.* # UI vermelha
│ └── icon-128.png
├── hub.mjs # bridge WS/HTTP
├── mcp.mjs # MCP stdio → hub
├── index.js # lib Playwright rápida
├── inject.js # mesmo runtime (lib)
├── exemplo.js
├── start-hub.bat
└── docs/
└── ARCHITECTURE.mdSecurity
Hub localhost only — don't expose the port
Extension asks for
debuggeronly for file upload (CDPDOM.setFileInputFiles)Don't run on shared machines without understanding what the agent can click
Use at your own risk; respect the ToS of the sites you automate
License
This server cannot be deployed
Maintenance
Related MCP Connectors
Hosted real Google Chrome MCP with per-user persistent state. Navigate, click, type, screenshot.
Browser MCP for logged-in tasks. Uses your Chrome — credentials stay local. Zero-token replay.
Automate cloud Chrome—navigate, click, type, screenshot, run code, record screen video
Run multi-step tasks in a real Chrome browser: persistent environments, live view, human takeover.
Related MCP Servers
- FlicenseCqualityDmaintenanceEnables browser automation, including navigation, form filling, login with CAPTCHA handling, and element manipulation, using a Chrome-based MCP server.364-
- AlicenseNot gradedqualityBmaintenanceEnables MCP clients to drive a real, logged-in Chrome browser for web automation tasks like navigation, clicking, typing, and screenshotting.4 npm1MIT
- AlicenseNot gradedqualityDmaintenanceEnables browser automation over MCP using a real Chrome browser with existing profile, supporting real tabs, downloads, cookies, and RPA workflows.44 npmMIT
- FlicenseNot gradedqualityCmaintenanceEnables MCP clients to control a real local browser window for web automation tasks such as clicking, typing, scrolling, and taking screenshots.11 npm-