Leaderboard/Automation/skills
Last commit on July 20, 2026·Created on February 6, 2026

browser-act/skills

A library of reusable automation packages for structured web interaction and data extraction
Combined rank
#77
across all skills
In Automation
#8
category rank
Stars
6k
+1.0% in last 7d
Forks
297
+1.4% in last 7d
Watchers
15
0.0% in last 7d
Traction scoreGitHub stars can be faked, so popularity alone can be misleading. Traction Score looks for broader signs of recent attention, adoption, and active maintenance.
TL;DR

This collection provides specialized skill packages that enable AI agents to navigate complex websites and perform repeatable tasks. It transforms manual site exploration into codified scripts, allowing builders to automate interactions without repeating the discovery process.

The repository focuses on high-fidelity data retrieval, such as extracting comprehensive property details from Airbnb via GraphQL APIs and performing bulk record scraping across various web platforms.

WHO IT'S FOR
AI agent / automation builders
creating reusable browser skills for websites
Web data extraction engineers
scraping bulk records from complex websites
E-commerce sourcing specialists
extracting wholesale product data from 1688.com
Backend / infrastructure engineers
analyzing internal website APIs for automation
Repository contents

103 skill files

Compatible AgentsThe repository documents support for these agents. The skills may also work with other agents that can load SKILL.md files, but they may need some setup or small changes.

Use this skill collection

Follow the documented setup, then try a first task.

Install browser-act through an AI agent

Installs: The browser-act Skill and its BrowserAct CLI integration

Before you start
  • Python 3.12 or newer
  • uv package manager
  • Chrome/Chromium when using the chrome or chrome-direct browser types
  • A BrowserAct API key for stealth browsers, stealth-extract, managed dynamic proxy, or solve-captcha; basic chrome and chrome-direct browsers work without authentication

Platform: Windows, macOS, or Linux

  1. In an AI agent session, enter this author-provided installation request exactly as shown: `Install browser-act. Skill source: https://github.com/browser-act/skills/tree/main/browser-act . Verify it works after installation.`

  2. If using a feature that requires an API key, obtain a key through BrowserAct registration and polling, or configure an existing key using the documented authentication instructions.

browser-act/SKILL.md · Checked Sep 18, 2026docs/installation.md · Checked Sep 18, 2026

Give it something to do.

Suggested first task

Analyze one Amazon product

Uses amazon-competitor-analyzer

  • A 10-character Amazon ASIN replacing <ASIN>
  • A valid BrowserAct API key configured in BROWSERACT_API_KEY

Check the repository for this skill’s setup.

Use the amazon-competitor-analyzer skill to analyze this single Amazon ASIN: <ASIN>. Confirm that BROWSERACT_API_KEY is configured, then report the product title, price, rating, review count, important specifications, and a short list of competitive strengths and vulnerabilities. Treat <ASIN> as the product's 10-character Amazon ASIN.
Suggested first task

Extract one webpage as Markdown

Uses web-search-scraper-api-skill

  • One HTTP or HTTPS webpage URL replacing <TARGET_URL>
  • A valid BrowserAct API key configured in BROWSERACT_API_KEY

Check the repository for this skill’s setup.

Use the web-search-scraper-api-skill to extract the readable content of <TARGET_URL> as Markdown. Confirm that BROWSERACT_API_KEY is configured first, and return the extracted content without publishing or modifying the source page. Replace <TARGET_URL> with one HTTP or HTTPS webpage URL.
Suggested first task

Extract one YouTube transcript

Uses youtube-transcript-extractor-api-skill

  • One YouTube video URL replacing <YOUTUBE_VIDEO_URL>
  • A valid BrowserAct API key configured in BROWSERACT_API_KEY

Check the repository for this skill’s setup.

Use the youtube-transcript-extractor-api-skill to extract the transcript and metadata from <YOUTUBE_VIDEO_URL>. Confirm that BROWSERACT_API_KEY is configured first, then return the video title, publisher, channel link, like count, and complete transcript. Replace <YOUTUBE_VIDEO_URL> with the URL of one YouTube video.