Give an AI agent its own real browser over MCP tool calls - launch, navigate, click, fill, screenshot, extract text, run JS - with a kernel-level real-device fingerprint and a persistent profile, so the session stays logged in between runs and pages see one coherent device instead of a headless build. No Playwright or SDK code to write.
Install
npx skills add antibrow/anti-detect-browser-skills --skill browser-mcp-agentnpx skills add antibrow/anti-detect-browser-skills --skill browser-mcp-agent -a claude-code -gnpx skills add antibrow/anti-detect-browser-skills --skill browser-mcp-agent -a codex -ggemini skills install https://github.com/antibrow/anti-detect-browser-skills.git --path browser-mcp-agentgh skill install antibrow/anti-detect-browser-skills browser-mcp-agentgit clone --depth 1 https://github.com/antibrow/anti-detect-browser-skills /tmp/antibrow__anti-detect-browser-skills
mkdir -p ~/.claude/skills ~/.agents/skills
cp -R /tmp/antibrow__anti-detect-browser-skills/browser-mcp-agent ~/.claude/skills/browser-mcp-agent # Claude Code
cp -R /tmp/antibrow__anti-detect-browser-skills/browser-mcp-agent ~/.agents/skills/browser-mcp-agent # Codex, Cursor, Gemini CLI, Copilot, OpenCode