Camofox Browser
github.com/jo-inc/camofox-browser- Category
- AI Agents
- Rank
- No. 417Tools index
Previous survey · No. 412 ·
- Pricing
- Open Source
- Platform
- cli · web
- Type
- TOOL
- Builder
- @jo-inc
- GitHub
- 10.4k stars
- Latest release
- v1.14.0
- Date
About
Anti-detection headless browser automation server for AI agents, built on Camoufox with C++ level fingerprint spoofing. Provides a REST API with accessibility snapshots, element references, and built-in anti-bot detection bypassing.
What it does
Camofox Browser gives agents a compact way to open pages, read meaningful controls, and act on them using stable references. It maintains separate browsing state per user, supports authenticated sessions and proxies, and can return screenshots, downloads, images, captions, or selected structured values.
Why it's ranked here
This is a strong fit when ordinary browser automation gets fingerprinted or produces too much page data for an agent. The combination of compact snapshots, persistent isolated sessions, proxy rotation, interactive login, tracing, metrics, and documented HTTP endpoints covers considerably more operational ground than a thin browser wrapper.
What's good
Stable element references make click and typing steps easier to reproduce. Large snapshots support truncation and pagination. Cookies and local storage survive restarts per user. Operators also get request IDs, optional Prometheus metrics, trace archives, OpenAPI documentation, download capture, upload controls, and configurable session limits.
Tradeoffs
Local installation requires Node.js 22 or newer and downloads roughly 300 MB on first run. Global bearer protection only applies when an access key is configured, so deployment security needs deliberate setup. Crash and hang telemetry requires explicit disabling. Schema-based extraction reads named snapshot references rather than discovering fields automatically.
How to use it well
Use it as the browser execution layer for an agent that repeatedly visits guarded or authenticated sites. Take a fresh snapshot before acting, persist login state, configure access control, and enable tracing while stabilizing workflows. Pair it with an external planner or job system because it executes browsing operations rather than supplying agent reasoning or crawl scheduling.
Technical notes+
package.json defines an ES module package with Node >=22, npm binaries for the browser server and MCP server, an optional Model Context Protocol SDK, Jest scripts, and a postinstall binary fetch. server.js assembles the Express service, Playwright-backed Camoufox lifecycle, per-session tabs, request metrics, structured logging, downloads, images, tracing, extraction, plugins, and Fly replay routing. lib/config.js centralizes environment parsing and resource limits; lib/auth.js implements timing-safe bearer checks and optional global gating; lib/proxy.js provides round-robin and backconnect strategies; lib/openapi.js mounts the generated specification and documentation. Makefile fetches architecture-specific Linux Camoufox and yt-dlp artifacts before Docker builds.
Observed
- License
- MIT
- Primary language
- JavaScript, with TypeScript plugin sources and configuration
- Packaging
- Published as the @askjo/camofox-browser npm package; requires Node.js 22 or newer
- Interfaces
- HTTP API, command-line binaries, MCP server, and OpenClaw plugin
- Deployment surface
- Local npm execution, Docker, Fly.io, and Railway
- Architecture support
- Docker tooling supports x86_64 and aarch64; Windows instructions use Docker Desktop with WSL2
- Testing
- Jest unit, end-to-end, plugin, live, MCP, and debug test scripts are declared
Read from README.md, Makefile, package.json, server.js, lib/fly.js, lib/auth.js, lib/proxy.js, lib/config.js, lib/images.js, lib/macros.js, lib/sentry.js, lib/cookies.js, lib/extract.js, lib/metrics.js, lib/openapi.js.
What it can do
Automate browser interactions via REST API
HTTP requests with automation commands → Browser actions and responses
Spoof browser fingerprints at C++ level
Target fingerprint configuration → Modified browser identity and characteristics
Generate accessibility snapshots of web pages
Web page URL or loaded page → Accessibility tree structure and element data
Create stable element references for web automation
Web page elements → Persistent element identifiers
Bypass anti-bot detection systems
Protected websites with bot detection → Successful access to restricted content
Run headless browser automation for AI agents
AI agent automation scripts and commands → Executed browser tasks and extracted data
Tags
Tech Stack
Comments (0)
No comments yet
Editorially curated, with community endorsements as a secondary signal. Corrections welcome.