A powerful browser automation CLI designed for AI agents to interact with web pages, extract dynamic JS content, bypass interactive verification, and handle multi-account sessions locally.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install browser-act-skills
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install browser-act-skills 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 browser-act?
browser-act is a comprehensive browser automation CLI tailored for AI agents, enabling them to control a full browser engine seamlessly. It serves as an advanced WebFetch or curl replacement, capable of retrieving JavaScript-rendered content, managing multi-browser sessions, executing complex interactions (like form-filling and file uploads), and capturing network data (XHR/fetch/HAR) in parallel.
Built with developer convenience and privacy in mind, this tool is one of the highly capable Openclaw Skills that ensures all cookie data, browser profiles, and credentials remain entirely local. By implementing human-agent collaboration workflows and verification-assistance APIs, it bridges the gap between pure script automation and guided manual overrides.
browser-act Use Cases
- Fetching, viewing, and extracting fully rendered, interactive JavaScript-heavy web pages.
- Automating complex browser workflows such as filling out web forms, clicking buttons, and uploading files safely.
- Managing multiple authenticated login sessions and browser profiles in parallel without cross-contamination.
- Retrieving real-time network interactions, including XHR, fetch, and HAR response logs, for data auditing.
- Bypassing visual challenges and captchas through integrated verification assistance and manual overrides via headed mode.
How browser-act Works
- Skill Loading: The AI agent initializes the skill and downloads the usage guide using
browser-act get-skills coreto retrieve environment rules. - Browser Selection & Creation: The CLI identifies or provisions isolated browser profiles based on agent demands, requiring explicit user confirmation for profile generation.
- Navigation & Interaction: The agent directs the browser to load URLs, execute clicks, scroll, type, and interact with the DOM.
- Dynamic Data Extraction: Content, screenshots, and network response logs are extracted and returned to the agent.
- Interactive Verification: If a captcha or login verification is encountered, the tool initiates the verification assistance protocol or requests human-agent manual intervention to proceed safely.
browser-act Setup
To begin using browser-act with Openclaw Skills, follow these steps to install and prepare your local environment.
Prerequisites
- Python 3.12 or newer
uvpackage manager installed on your machine
Installation
Run the following command in your terminal to globally install the browser-act CLI:
uv tool install browser-act-cli --python 3.12
Operational Preparation
Before executing any automation commands, always initialize the tool by fetching the core usage workflows. Do not truncate this command's output, as it is critical for correct environment setup:
browser-act get-skills core --skill-version 2.0.2
browser-act Data Schema & Taxonomy
browser-act organizes its local runtime files and metadata structure within a dedicated CLI data directory to maintain strict boundaries. Below is the file and data structure taxonomy:
| Directory/File Path | Type | Description |
|---|---|---|
CLI Data Directory |
System Path | Root folder containing all browser profiles, logs, and configuration states. |
/profiles/ |
Subdirectory | Isolated browser profile directories storing cookies, login session states, and local storage per-browser. |
/logs/ |
Subdirectory | Diagnostic session logs rotated automatically after each CLI run execution. |
Captchas / Verification Images |
Cache | Temporary local files containing captcha challenge images, only transmitted externally if captcha solver is explicitly invoked. |
All cookies and browser profiles are locked to local storage and are never uploaded, ensuring maximum privacy compliance under standard Openclaw Skills data guidelines.
browser-act Advanced Features
- Safety Confirmation Gate: Implements safety protocols requiring explicit user approval before browser creation, deletion, or performing sensitive operations like form submissions.
- Headed Mode & Remote Assist: Allows seamless human-agent collaboration by spawning a visual browser UI or providing remote assist mechanisms for handling complex manual workflows.
- Verification Assistance API: Out-of-the-box support for sending visual challenge images to verification endpoints when user-authorized captcha solving is triggered.
- CDP Connection: Supports direct Chrome DevTools Protocol (CDP) connection to local Chrome instances, granting advanced raw browser orchestration capabilities with user confirmation.
SKILL.md
Loading
Related Openclaw Skills
A2A Chat - Agent-to-Agent Realtime Communication
A2A Chat is a lightweight, real-time messaging protocol that enables multiple AI agents to dynamically create chat rooms, collaborate, and exchange data seamlessly.
Browser Act Skill Forge
An AI agent utility that transforms interactive browser-based web exploration, scraping, and action sequences into permanently reusable Openclaw Skills.
YouTube Transcript Extraction & Content Reformatting
Automate fetching YouTube transcripts to generate timestamped chapters, blog posts, X/Twitter threads, or summaries instantly.
Instagram Post Comments
An automated browser-driven skill to extract comments, user handles, timestamps, and reaction metrics from any Instagram post using cursor-based pagination.
Zhihu Search & Content Extractor
An API-driven, token-optimized content extraction skill designed to allow AI agents to search, retrieve, and summarize high-quality Zhihu content without rendering a heavy browser.
API Document Generator
An automated engineering skill that converts raw source code, URLs, or endpoint descriptions into enterprise-ready OpenAPI 3.0 standards, complete with multi-language SDKs and mock servers.
Featured*
