Tencent Open-Sources BrowserSkill to Connect AI Agents to Real Browsers
Tencent Open-Sources BrowserSkill to Connect AI Agents to Real Browsers
Tencent released BrowserSkill, an open-source tool that allows AI agents to interact with the user's actual browser session, leveraging existing logins and cookies.
Tencent has open-sourced BrowserSkill, a tool that connects AI agents directly to your real browser environment, rather than forcing them to operate in isolated, clean sessions.
Typically, an agent gets a separate browser instance without your accounts. BrowserSkill enables agents like Codex, Claude Code, Cursor, and others to operate on websites where you are already authenticated.
The agent acts within a separate, visible window. If it needs access to your specific tab, it first requests permission, temporarily takes control, and returns it after completing the task. Captchas and confirmation dialogs are handed over to the user, after which the workflow continues.
This is not an MCP server, but a local command-line utility and browser extension. Therefore, BrowserSkill works with any agent capable of executing terminal commands.
It supports Windows, macOS, and Linux. The entire code is distributed under the MIT license.
Why it matters
AnalysisThis bridge between terminal-based agents and authenticated browser sessions unlocks new automation workflows, allowing AI to perform tasks that require user login states while maintaining security through explicit permission requests.
Discuss in community
Share your questions and insights with developers
Related in Ecosystem
Practical guides, core terminology, and agent skills
AGENTS.md in Ghostty and Turborepo makes AI agents report on users
Examples of AGENTS.md files show how repository instructions can force AI agents to add insulting notes about developers to pull requests.
Jev and System One AI: A Complete Practical Guide to the Decision Model from TypeSafe
A deep engineering breakdown of the flagship Jev model: System One architecture versus LLMs, Choice/Score/Noul primitives, request parallelism, confidence mathematics, video use cases, code, and an overview of local open-weight alternatives (Laya, GLiNER2.5, CLM-8B).
DeepSeek launches Creator Mode to let users reshape their harness
DeepSeek’s open-source harness now supports Creator Mode, where an agent can write and embed plugins based on plain-language descriptions.
Claude Code Projects: How a Single AI Orchestrates a Team of Autonomous Agents
A deep dive into Claude Code Projects: Coordinator and Threads architecture, parallel branches, shared memory via MEMORY.md, and cloud agent management.