camofox-browser - Open-Source AI Agent Anti-Detection Browser

What is camofox-browser

Camofox Browser is an open-source AI Agent anti-detection browser server built on Camoufox (a Firefox modified at the C++ level for fingerprint spoofing). It provides AI with structured webpage snapshots via a REST API (over 90% size reduction), generating stable element references (such as e1, e2) for AI to interact with directly. Camofox Browser has built-in C++-level fingerprint spoofing, capable of bypassing mainstream anti-scraping verification such as Cloudflare; it supports Cookie import, session persistence, and 15+ search macros.

Main features of camofox-browse

• C++-level anti-detection browsing: Built on Camoufox, a modified Firefox, it spoofs fingerprint information such as hardware concurrency, WebGL renderer, AudioContext, screen geometry, and WebRTC at the C++ implementation layer, which cannot be seen through at the JavaScript layer.
• Structured webpage snapshots: Distills raw HTML into a numbered, labeled accessibility snapshot (over 90% size reduction), allowing AI to interact with the page directly through e1、e2 and other stable references.
• Cookie import and session persistence: Supports importing Netscape-format Cookie files for login-free access; user session Cookies and localStorage are automatically persisted locally, maintaining login status after restart.
• Search macro shortcuts: Built-in 15+ search macros for mainstream platforms (such as @google_search、@reddit_search、@youtube_search), reaching the target page with a single command; the Reddit macro returns JSON data directly.
• Automatic proxy and GeoIP matching: Supports residential proxies and Backconnect rotating proxies, automatically setting the browser timezone, language, and geographic coordinates based on the proxy exit IP to maintain fingerprint consistency.
• YouTube subtitle extraction: Integrates yt-dlp to quickly extract subtitles from any YouTube video, no API Key required, with multi-language filtering support.
• VNC visual interactive login: Manually complete interactive login on complex websites in the browser via noVNC, and export the storage state for reuse by subsequent Agent automation.
• Session tracing and debugging: Supports enabling Playwright tracing per session, generating .zip files containing screenshots, DOM snapshots, and network requests for later troubleshooting.
• Lightweight deployment and elastic scaling: Memory usage is about 40MB when idle, supports one-click deployment on Docker, Fly.io, and Railway, with lazy loading of the browser engine + automatic shutdown when idle.

Technical principles of camofox-browse

• C++-level fingerprint spoofing: Camoufox modifies at the Firefox source code level navigator.hardwareConcurrency, WebGL render strings, AudioContext fingerprints, screen resolution, and WebRTC and other attributes, so anti-scraping systems receive spoofed parameters before JS execution.
• Accessibility Tree snapshots: Uses Playwright's accessibility tree to extract the page's semantic structure, generating a textual snapshot with stable references to pass to the LLM instead of the raw DOM.
• Playwright browser context isolation: Each user corresponds to an independent BrowserContext, with Cookies, localStorage, and cache fully isolated to avoid session cross-talk.
• Lazy loading and idle recycling: The Camoufox engine (about 300MB) is started on the first request, and the process automatically shuts down after 5 minutes with no active sessions, reducing resident resource usage.
• Structured JSON logs: All request and response outputs are single-line JSON objects, including reqId for log aggregation and traceability in distributed environments.

How to use camofox-browse

• Install dependencies: Run npm install @askjo/camofox-browser. On first installation, it will automatically download the approximately 300MB Camoufox engine.
• Local startup: After cloning the repository, run npm install && npm start. The service listens by default on http://localhost:9377。
• Configure access key (production environment): Generate a 32-character hexadecimal key and set the environment variable CAMOFOX_ACCESS_KEY. All requests must carry Authorization: Bearer <key>。
• Import cookies to achieve login-free access: Place the Netscape-format cookie file exported from the browser into the ~/.camofox/cookies/ directory, and use the API or the OpenClaw plugin to call the import tool to inject the session.
• Create a tab and get a snapshot: Send a POST request to /tabs to create a tab, then call /tabs/:id/snapshot to get a e1、e2 referenced structured page summary.
• Perform page interactions: Through /tabs/:id/click or /tabs/:id/type , pass in the element reference number and text content to implement click and input operations.
• Navigate using search macros: Call /tabs/:id/navigate and pass in "macro": "@google_search" and "query": "关键词"to directly get the search results page.
• Docker deployment: In the project root directory, execute make up. The Makefile will automatically detect the CPU architecture and pre-download the binary files, completing the build and startup in about 30 seconds.

Core advantages of camofox-browse

• Bypass mainstream anti-scraping and CAPTCHAs: Because fingerprints are forged at the C++ layer, it can stably pass bot detection on platforms such as Google and Cloudflare, avoiding frequent triggering of human verification.
• Extremely high token efficiency: Structured snapshots reduce the size by more than 90% compared with raw HTML, greatly reducing LLM context window usage and API call costs.
• Stable and reliable element references: After minor page layout adjustments, references such ase1、e2 can still remain valid through the accessibility tree, reducing the probability of Agent operation failures due to DOM changes.
• Out-of-the-box search capabilities: 15+ search macros cover mainstream information sources, so the Agent does not need to construct search URLs and parse pages itself. Sites such as Reddit directly return structured JSON.
• Production-grade security model: Cookie import is disabled by default (requires CAMOFOX_API_KEY), supports global Access Key authentication, path traversal protection, runtime injection of all sensitive data, and zero hard-coded keys.

Project address of camofox-browse

• GitHub repository：https://github.com/jo-inc/camofox-browser

Comparison of camofox-browse with similar competitors

Use cases for camofox-browse

• AI Agent web automation: Provides a programmable browser backend for LLM-based agents to perform tasks such as data scraping, form filling, and information retrieval.
• E-commerce and competitor monitoring: Bypasses platform anti-scraping mechanisms to automatically scrape product prices, inventory, and review information, supporting account-level data collection under logged-in sessions.
• Social media automation: After importing Cookies, automatically operates accounts such as Twitter, LinkedIn, and Instagram to perform content publishing, interaction, and data analysis.
• Academic and public opinion research: Uses search macros to quickly retrieve information from Google, Wikipedia, and Reddit, and extracts YouTube video subtitles for content analysis.
• Cloud automated testing: Deployed in VPS or container environments, serving as a headless browser service to support end-to-end testing and web monitoring in CI/CD pipelines.