OyadotAI/oya-browser

The browser control plane for AI agents. One API over Oya Cloud, Browserbase, Steel, Anchor, Browser Use and your own Chrome, with persistent personas, CAPTCHA and MFA handling, and live human takeover.

What it solves

Oya Browser addresses the fragility and detectability of traditional browser automation. Most automation tools drive browsers from the outside via debugging protocols, which are easily flagged as bots and often struggle with complex authentication (SSO, passkeys, Cloudflare). Oya integrates automation directly into the browser process, making agents appear as real users and simplifying session management.

How it works

Oya is a custom browser where automation logic lives inside the process rather than being attached externally. It uses Chrome's own emulation for platform and locale settings to avoid bot detection. For AI agents, it provides an SDK and an MCP (Model Context Protocol) interface, allowing LLMs to navigate pages and extract data using plain-language commands. It also features a "record and replay" system where a model-driven task can be saved as a "playbook" (a Playwright module) and re-executed without needing an LLM in the loop for subsequent runs.

Who it’s for

  • AI Agent Developers: Those building agents (like Claude Code or Cursor) that need a reliable, stealthy way to interact with the web.
  • Automation Engineers: Users who want to record complex web workflows and replay them efficiently without recurring model costs.
  • Enterprises: Organizations requiring an audit trail of browser actions and the ability to self-host their browser infrastructure on Docker, Kubernetes, or AWS ECS.

Highlights

  • Stealth Architecture: Achieves a 0% CreepJS headless score by running automation inside the browser process.
  • Session Persistence: Imports logins from existing browsers (Chrome, Arc, etc.) so agents start already signed in.
  • Playbook System: Records AI-driven actions into reusable scripts that run without a model for repeated tasks.
  • Agent-First Integration: Supports MCP for seamless integration with tools like Claude Desktop and Windsurf.
  • Flexible Deployment: Can be self-hosted via Docker or deployed to managed cloud environments.

Related

  • Project
    anomalyco/browser-control

    Browser Control is an open‑source Node.js driver that lets AI agents or CLI users run Playwright code inside their own Chromium browser (with all logged‑in sessions and extensions). It works via a local relay and a browser extension, supports persistent sessions, inspection helpers, human hand‑offs, video/ HAR recording, and a TypeScript Effect client for secret‑safe API calls. Intended for trusted local use only.

  • Project
    remorses/playwriter

    A browser automation tool that lets AI agents control your existing Chrome browser session, maintaining logins and extensions to avoid bot detection.

  • Project
    browseros-ai/BrowserOS

    BrowserOS is an open‑source Chromium‑based browser suite that lets AI agents (Claude Code, Codex, Cursor, etc.) drive real‑world web tasks using your own logged‑in accounts. It offers a live dashboard, session replay, parallel agents, and full‑device privacy, with installers for macOS and Windows.

  • Project
    sleepinginsummer/agent-browser-cli

    A CLI tool that connects AI Agents to a real Chrome browser session, allowing them to reuse login states and perform precise page control and perception.

  • Project
    Player-YN/BrowserKitten

    PawWork · 爪爪 is a Chrome MV3 extension that adds a side‑panel AI agent. You supply an OpenAI API key, then tell the agent in natural language to control the current tab, run JavaScript, or store small data artefacts. It’s built with a service‑worker → off‑screen worker → ToolLoopAgent architecture, requires Chrome 135+, and is intended for users who want LLM‑driven browser automation without writing code.