AI Agent External Interaction

use-my-browser

by @xixu-me

Intelligent browser automation strategy that automatically selects the best solution based on task type: public web tools, real-time browser sessions, or independent browsers.

4.8(1.7K)

153.7K

dom-manipulation

browser-use

by @browser-use

A front-end development capability designed for AI coding agents, focusing on browser usage. It enables AI agents to understand and manipulate web elements, perform front-end tasks, and thus play a role in web interface development.

4.7(2.0K)

76.1K

mcp-builder

by @anthropics

Guides the creation of high-quality MCP (Model Context Protocol) servers, enabling Large Language Models (LLMs) to interact with external services through well-designed tools.

4.7(1.1K)

59.5K

api

better-auth-best-practices

by @better-auth

Provides Better Auth integration guide and best practices, instructing users on installation, environment variable configuration, and referring to official documentation for the latest API and code examples.

4.7(1.3K)

52.7K

firecrawl-scrape

by @firecrawl

Extract clean Markdown content from any URL, including JavaScript-rendered SPAs, for web content retrieval.

4.7(532)

44.4K

web-search

firecrawl-search

by @firecrawl

Performs web searches and extracts full page content, providing comprehensive support for data analysis and information retrieval, obtaining in-depth information.

4.7(112)

44.2K

ai-agents

firecrawl-agent

by @firecrawl

AI-driven autonomous data extraction tool capable of navigating complex websites and returning structured JSON data.

4.7(108)

43.5K

playwright-cli

by @microsoft

Automate browser interactions for web testing, form filling, screenshots, and data extraction. (AI Agent Skill for enhanced efficiency and automation)

4.6(588)

39.0K

notion-integration

notion-api

by @intellectronica

Provides a comprehensive guide for interacting with the Notion API, helping users integrate and automate Notion workflows for data synchronization and management.

4.6(108)

30.3K

data-extraction

firecrawl-interact

by @firecrawl

Interacts with crawled pages in a real-time browser session, supporting clicks, form filling, navigation, and data extraction.

4.6(25)

28.7K

firecrawl-build-scrape

by @firecrawl

Extracts full webpage content from known URLs, supporting retrieval augmentation, summary generation, and content monitoring.

4.6(11)

26.9K

firecrawl-build-interact

by @firecrawl

Crawls dynamic content through interactive operations like clicks, inputs, and navigation, supporting multi-step processes, form submissions, and pagination.

4.6(7)

26.8K

google-sheets-api

gws-sheets-read

by @googleworkspace

Read values from specified cells or ranges in Google Sheets spreadsheets, facilitating data retrieval.

4.6(309)

24.9K

gmail-api

gws-gmail-send

by @googleworkspace

Send emails via Gmail, supporting specified recipients, subject, and email content.

4.6(298)

23.3K

anti-detection

anti-detect-browser

by @antibrow

Launch Chromium instances with real device fingerprints via Playwright API to bypass anti-bot system detection.

15.6K

google-drive

recipe-bulk-download-folder

by @googleworkspace

Lists all files in a specified Google Drive folder and supports bulk downloading.

15.1K

google-drive-api

recipe-watch-drive-changes

by @googleworkspace

Subscribe to change notifications for Google Drive files or folders to stay updated on file modifications in real-time.

14.8K

news-headlines

the-news

by @sfkislev

This skill provides AI agents with real-time and historical access to front-page headlines from newspapers and news sites across 20 countries via a public API. It delivers main headlines with AI-generated overviews for context, enabling agents to quickly grasp breaking news, current events, and perform comparative media analysis. The skill offers a timestamped, multi-source snapshot for efficient global perspective acquisition, news analysis, and historical data retrieval.

13.3K

flue

by @sfkislev

Flue is a local automation layer that enables AI agents to control desktop software like Adobe Photoshop, Blender, Unity, and Microsoft Office. It executes in-application scripts via bridge commands and returns structured JSON, achieving deep interaction and automated operations with software.

12.2K

firecrawl-browser

by @firecrawl

Provides cloud-based browser automation services to handle interactive web pages, enabling data scraping and content extraction, improving efficiency.

4.5(112)

11.3K

microsoft-documentation

microsoft-docs

by @github

Query official Microsoft documentation for concepts, tutorials, and code examples related to Azure, .NET, Agent Framework, VS Code, and more.

4.5(319)

10.9K

cli-utilities

opencli-usage

by @jackwener

OpenCLI transforms any website into a CLI for humans and AI agents, enabling browser automation like navigation, clicks, and data extraction. It also unifies local tools and Electron apps into a single CLI hub.

10.9K

youtube

baoyu-youtube-transcript

by @jimliu

Automatically download YouTube video subtitles and transcripts, supporting manual and auto-generated captions, without API keys or browser, directly calling YouTube InnerTube API.

4.5(25)

10.6K

smart-search

by @jackwener

OpenCLI converts any website into a command-line interface, automating operations via your logged-in Chrome browser. It empowers AI agents to navigate, fill forms, click, extract information, and automate web interactions. Additionally, it serves as a hub for local CLI tools, supporting the integration of existing command-line utilities and desktop applications, significantly boosting the efficiency of both human and AI agents.

9.6K

nano-banana-pro

nano-banana-pro-openrouter

by @github

Generates or edits images using the Gemini 3 Pro image model via OpenRouter, supporting pure prompt generation, image editing, and multi-image composition.

4.5(309)

9.5K

notebooklm

notebooklm-py

by @teng-lin

notebooklm-py is an unofficial Python API and AI skill library for Google NotebookLM, offering complete programmatic access, including features not exposed in the web version. It allows users to automate operations via Python, CLI, or AI agents (such as Claude Code), easily create notebooks, add various sources (e.g., URLs, PDFs, videos), interact with content, generate various documents and content, and supports downloading in multiple formats, significantly boosting research and content creation efficiency.

4.5(379)

9.5K

mcp-cli

by @github

Interact with MCP servers via CLI to connect external tools, APIs, or data sources, and list available resources.

4.5(308)

9.2K

stripe-api

stripe-integration

by @wshobson

Focuses on backend Stripe payment integration, combining intelligent automation and multi-agent orchestration to simplify payment processes and improve transaction efficiency.

4.5(90)

9.0K

opencli-browser

by @jackwener

Controls Chrome browser via command line, reusing existing login sessions for passwordless browser automation.

4.5(3)

9.0K

opencli-autofix

by @jackwener

Automatically diagnoses and fixes adapters when website DOM or API changes cause OpenCLI command failures, requiring no manual intervention.

4.5(13)

8.7K

web-access

by @eze-is

Access and interact with web content via Chrome DevTools Protocol, supporting automated browser operations, data scraping, and form filling.

8.5K

microsoft-power-platform

power-platform-mcp-connector-suite

by @github

Generate a complete Power Platform custom connector, integrating MCP to support Copilot Studio, including schema generation and troubleshooting.

4.5(289)

8.4K

typespec

typespec-create-api-plugin

by @github

Generate a TypeSpec API plugin for Microsoft 365 Copilot, including REST operations, authentication, and Adaptive Cards.

4.5(289)

8.4K

slack-api

slack

by @vercel-labs

Interact with Slack workspaces via browser automation to check unread channels, send messages, extract data, etc.

4.5(297)

7.4K

opencli-adapter-author

by @jackwener

This skill guides users through the end-to-end process of writing an OpenCLI adapter for a new site or adding a new command to an existing one. It covers initial reconnaissance, API discovery, field decoding, adapter coding, and verification, aiming for a complete cycle from scratch to passing `opencli browser verify` within 30 minutes.

6.8K

web-monitoring

parallel-monitor

by @parallel-web

This skill enables AI agents to continuously monitor web content for changes, such as price updates, news, or document filings. It supports creating, listing, updating, and deleting monitoring tasks, and provides notifications when changes are detected, facilitating automated data tracking and real-time information retrieval.

6.6K

claude-api

claude-to-im

by @op7418

Manages the bridging of Claude with Instant Messaging (IM) applications, handling user data storage and skill directories to enable Claude-IM interaction.

4.4(20)

5.9K

by @sickn33

Browser automation techniques supporting web testing, data scraping, and AI agent interaction to improve efficiency.

4.4(84)

4.9K

ai-agents

agent-reach

by @panniantong

This skill is a guide to using Agent Reach, providing upstream tools for 13+ platforms, allowing direct calls and checking available channels.

4.4(38)

4.9K

ai-engineering

instaclaw

by @napoleond

An image sharing platform designed for AI agents, utilizing ATXP authentication for secure and efficient image exchange and collaboration among agents.

4.4(51)

4.5K

ai-agent-development

clawdirect-dev

by @napoleond

Follows ClawDirect pattern to build proxy-oriented ATXP authenticated web experiences for AI agent interaction website development.

4.4(171)

4.4K

api-development

shopify-storefront-graphql

by @shopify

Builds custom Shopify storefronts using the Storefront GraphQL API, querying product, cart, and checkout data.

4.4(8)

4.3K