osaurus-browser

Automate headless browser interactions with per-agent session persistence and secure sign-in flows.

54|9|Updated Nov 22, 2025
One-click install
npx skills add https://github.com/osaurus-ai/osaurus-tools --skill osaurus-browser
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: osaurus-browser
Source: https://github.com/osaurus-ai/osaurus-tools/tree/main/tools/browser
Command: npx skills add https://github.com/osaurus-ai/osaurus-tools --skill osaurus-browser

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Osaurus Browser provides headless browser automation with per-agent session persistence, enabling agents to navigate sites, sign in via a dedicated helper window, and collect page snapshots without exposing credentials in chat, while keeping sessions isolated between agents.

Core Features & Use Cases

  • Per-agent persistent browser sessions with on-disk storage for cookies, localStorage, and IndexedDB.
  • Sign-in helper window (browser_open_login) to securely authenticate without sharing credentials.
  • Element refs and detail levels to drive interactions and inspect results; built-in inspection tools for console and network requests; multi-agent coordination via browser_lock.

Quick Start

Open a login window when required, then navigate and interact with pages using browser_navigate and browser_do while preserving per-agent browser sessions.

Frequently Asked Questions about osaurus-browser

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate headless browser sign-in without exposing credentials in chat?

Headless browser sign-in can be automated securely using a dedicated helper window that authenticates per-agent sessions without exposing credentials in chat. This isolates login flows while persisting cookies and localStorage on disk.

What is per-agent session persistence in headless browser automation?

Per-agent session persistence isolates browser sessions on disk, storing cookies, localStorage, and IndexedDB separately for each agent. This ensures multi-agent environments maintain independent authenticated states across navigations.

How do I keep browser sessions isolated when running multiple agents?

Browser sessions are kept isolated across multiple agents using per-agent on-disk storage and a browser lock mechanism for coordination. This separates cookies, localStorage, and IndexedDB so agents never share session state.

Can I inspect console and network requests during headless browser automation?

Console and network requests can be inspected during headless browser automation using built-in inspection tools. These tools capture request details and console output while agents navigate pages and drive interactions.

Does headless browser automation work for multi-agent data gathering tasks?

Headless browser automation works for multi-agent data gathering by combining per-agent session isolation with page snapshot collection. Agents navigate sites, sign in securely, and extract data without sharing credentials or session state.

What do I need to start automating browser interactions with per-agent isolation?

To start automating browser interactions with per-agent isolation, you need three core functions: browser_open_login to authenticate, browser_navigate to load pages, and browser_do to drive interactions. These maintain persistent sessions across tasks.