Smooth Browser for Openclaw

An AI-powered browser tool for agents to navigate websites, handle authentication, and extract structured data using natural language commands.

antoniocirclemind
v1.0.0
Feb 6, 2026
0
2.4k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install smoothbrowser

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install smoothbrowser using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Smooth Browser?

Smooth Browser is a specialized browser environment built for AI agents to interact with the modern web as if they were human users. It provides a CLI-based interface that allows agents to carry out complex workflows such as navigating websites, filling out forms, and extracting data from dynamically rendered pages. By integrating this capability into the Openclaw Skills framework, developers can enable their agents to perform autonomous research, test web applications, and automate repetitive online tasks with ease.

Unlike traditional automation tools that require brittle CSS selectors, Smooth Browser uses natural language processing to understand goals. It features robust profile management to persist cookies and login sessions, ensuring that agents can maintain authenticated states across multiple sessions. Whether it is for scraping real-time data or performing multi-step transactions, this skill provides the necessary infrastructure for reliable web-based agentic behavior.

Smooth Browser Use Cases

  • Persistent authentication for websites requiring logins or 2FA via session profiles.
  • High-fidelity web scraping that transforms unstructured web content into clean JSON data.
  • Automated testing of web applications using natural language user stories.
  • Complex data entry and form submission across enterprise web portals.
  • Real-time monitoring of web content with human-in-the-loop intervention capabilities.

How Smooth Browser Works

  1. Initialize a persistent profile to store cookies, sessions, and browser state for future use.
  2. Start a browser session, optionally specifying device types, allowed URLs, or associated file assets.
  3. Send natural language tasks to the agent using the run command, which executes the browser actions autonomously.
  4. Use the extract command for high-efficiency data collection when specific structured output is required from a page.
  5. Intervene manually through a live-view URL if the agent encounters challenges like CAPTCHAs or complex UI patterns.
  6. Close the session to ensure all state changes are saved to the profile and retrieve any downloaded files.

Smooth Browser Setup

To get started with this skill within the Openclaw Skills environment, follow these installation steps:

# Install the Smooth CLI
pip install smooth-py

# Configure your API key (get one at app.smooth.sh)
smooth config --api-key <your-api-key>

# Verify your configuration
smooth config --show

Smooth Browser Data Schema & Taxonomy

Smooth Browser organizes its operations around sessions and profiles to ensure data integrity and persistence. Below is the primary data structure used by the skill:

Entity Data Type Purpose
Profile Persistent Store Maintains cookies, local storage, and session state for specific accounts.
Session Active Process An ephemeral or profile-based browser instance identified by a unique ID.
File Binary Asset Uploaded documents (PDFs, images) available for the agent to use in-session.
Structured Output JSON Schema User-defined metadata templates for consistent data extraction results.

Smooth Browser Advanced Features

  • Profile Persistence: Reuse browser states and cookies across different sessions to bypass repetitive logins.
  • Structured Data Extraction: Define custom JSON schemas for the agent to return perfectly formatted data.
  • Multi-Agent Coordination: Support for concurrent sessions, allowing subagents to work on independent web tasks simultaneously.
  • Live View Interface: Provides a real-time URL for developers to watch or interact with the agent's browser window.
  • Integrated Proxy Management: Automatic handling of built-in proxies to prevent IP blocking and improve reliability.
  • Custom JavaScript Execution: Ability to evaluate complex JS snippets directly within the browser context for advanced DOM manipulation.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*