HeadlessX is a self-hosted browser automation platform for scraping and agent workflows. It pairs a web dashboard and protected API with queue-backed jobs and an MCP endpoint so coding agents can drive stealth browser sessions without a managed Browserless bill.
Key Features
- Stealth Stack: Built around Camoufox/Firefox-oriented anti-detection approaches.
- Dashboard + API: Operate jobs from UI or integrate via HTTP.
- Queues: Background workflows instead of fragile one-off scripts.
- MCP Endpoint: Hook agents (Claude Code, Cursor, etc.) into browser tools.
- Operator Toolkit: Scrape, search, and extract patterns for AI pipelines.
- Self-Hosted: Run on your VPS; MIT licensed.
Why Choose HeadlessX?
- Cloud browser APIs get expensive at volume.
- You need automation that agents can call over MCP.
- Puppeteer-only scripts lack a durable ops UI.
- Privacy-sensitive scrapes should stay on your network.
Docker Deployment
git clone https://github.com/saifyxpro/HeadlessX.git
cd HeadlessX
# follow README for compose / env (API keys, browser deps)
docker compose up -dCheck the repo for current image names and resource recommendations (browsers need RAM).
Getting Started
- Deploy HeadlessX and secure the API.
- Run a sample scrape from the dashboard.
- Connect MCP from your agent client.
- Build queued jobs for recurring extracts.
- Point results into n8n, databases, or RAG stores.
Full Setup Guide
github.com/saifyxpro/HeadlessX. Combine with Executor.sh MCP gateway and Dokploy for production hosting.
