AI Briefing#browser-automation
The latest AI and developer news about #browser-automation, with the original source and a short summary.
The latest AI and developer news about #browser-automation, with the original source and a short summary.
The stateless browser runs on Cloudflare Workers and supports Chrome DevTools Protocol for agent interaction.
The feature allows the agent to run multiple tasks at once while users control login choices and approve downloads.
The update introduces a sandboxed in-app browser, rebuilt computer use controls, and a 40% reduction in pre-message context size.
The update introduces moss-image generation, parallel tool calls, and LLM tuning alongside improved voice cloning and computer vision capabilities.
The MIT-licensed tool enables AI agents like Cursor and Claude Code to automate tasks using the user's existing Chrome or Edge sessions via a dedicated CLI and browser extension.
Stagehand v4 supports TypeScript, Python, and Go, delivering execution speeds 2x faster than Playwright in the Browserbase cloud environment.
BrowserSkill, a CLI tool for agents that reuses actual logged-in browser sessions while operating in a separate window, has been released.
Open-source tools Savage MCP and Savage Scraper have been released to minimize browser control permissions for LLM agents.
OpenClaw v2026.9.3 has been released, adding key features such as enhanced update recovery, real-time browser monitoring, and cloud repository execution.
Microsoft and Google have proposed 'WebMCP', a web standard that enables AI agents to directly recognize web page capabilities.
Claude Cowork can now perform web tasks using its own browser without requiring a Chrome extension.
Anthropic officially launched its computer use and skills APIs, enhancing efficiency for production agents.
Claude has introduced full-fledged collaboration sessions that perform tasks by directly manipulating web pages through the Chrome side panel.
Cloudflare has built Kitesurf, a browser dedicated to AI agents, on top of Workers.
The agent-browser extension now lets the eve agent navigate and manipulate the web.
Microsoft Research has released Fara1.5-27B, a multimodal web browser automation agent based on Qwen3.5-27B, on Hugging Face.
Running web agents at production scale requires five layers beyond the browser itself: warm pools, isolation, identity, observability, and a model gateway.
Socar's Modu-ui-Jucharjang built a LangGraph-based AI agent system that automatically recovers from discount-application failures across 50+ external equipment vendor websites.
Moonshot AI has launched Kimi Work, a desktop AI agent that supports local file integration, web automation, and 24-hour scheduling.
It's a tool that converts a webpage's clickable and input elements into structured JSON to help AI agents with browser automation.
Mozilla has launched Tabstack, an API that extracts web data into structured JSON and automates browsers.
LangChain overcomes the execution environment limitations of agents through Headless Tools, which can directly access client-side APIs and state.
Intuned has launched as a platform for building and maintaining browser automation with code through AI agents.
Microsoft has unveiled MagenticLite and Fara1.5, agentic systems designed for small models.