Home
Welcome to the Agentic Web Learning Tool wiki!
The Agentic Web Learning Tool is a Chrome Extension designed to serve two primary purposes:
- Educational Platform: To help developers and researchers understand the concepts behind the Agentic Web, including the Model Context Protocol (MCP) and WebMCP.
- Functional Development Environment: To provide a robust set of tools for building, debugging, and testing agentic browsing solutions.
- Developers looking to build AI agents that can interact with the web.
- Researchers exploring the intersection of LLMs and browser automation.
- Learners who want to understand how standard interfaces (MCP) can bridge the gap between AI models and browser contexts.
- Side Panel Chat Interface: Interact with LLMs directly within the browser, utilizing context from the current page.
- MCP Server Management: Connect to and manage external MCP servers to extend your agent's capabilities.
- WebMCP & Chrome API Tools: A browser-side implementation that exposes DOM elements as tools to the agent, plus Chrome API tools for tab management, history, scripting, storage, and more — including experimental cross-document tool execution.
- Built-in AI: Leverage Chrome's on-device Gemini Nano model through the Prompt, Writer, Rewriter, Summarizer, Proofreader, Translator, and Language Detector APIs — with interactive playgrounds for each.
- Multi-Provider Support: Connect to Browser AI (Gemini Nano), Anthropic, OpenAI, or Google Gemini — with Extended Thinking support for compatible models.
- Workflow Engine: Create and execute automated workflows with a visual node-based editor, including built-in demo workflows and support for DOM manipulation, loops, and all Built-in AI APIs.
- MCP Inspector & DevTools: Debug and inspect the state of your agents, tools, and breakpoints directly in Chrome DevTools.
- Getting Started: Installation and setup instructions.
- Evaluation Environment: Setting up Chrome with Built-in AI flags.
- Core Concepts: Deep dive into Agentic Web, MCP, WebMCP, and Built-in AI.
- Configuration: Setting up LLM providers and options.
- Features: Detailed guides on Built-in AI, MCP Servers, WebMCP, Workflows, and MCP Inspector.