MCP Server for the RAG Web Browser Actor 🌐

1# MCP Server for the RAG Web Browser Actor 🌐
2 
3Implementation of an MCP server for the [RAG Web Browser Actor](https://apify.com/apify/rag-web-browser).
4This Actor serves as a web browser for large language models (LLMs) and RAG pipelines, similar to a web search in ChatGPT.
5 
6> **This MCP server is deprecated in favor of [mcp.apify.com](https://mcp.apify.com)**
7>
8> For the same functionality and much more, please use one of these alternatives:
9 
10## 🚀 Recommended: use mcp.apify.com
11 
12The easiest way to get the same web browsing capabilities is to use **[mcp.apify.com](https://mcp.apify.com)** with default settings.
13 
14**Benefits:**
15- ✅ No local setup required
16- ✅ Always up-to-date
17- ✅ Access to 6,000+ Apify Actors (including RAG Web Browser)
18- ✅ OAuth support for easy connection
19- ✅ Dynamic tool discovery
20 
21**Quick Setup:**
221. Go to https://mcp.apify.com
232. Authorize the client (Claude, VS Code, etc.)
243. Copy the generated MCP server configuration (or use OAuth flow if supported)
254. Start using browsing & other tools immediately
26 
27## 🌐 Alternative: direct RAG Web Browser integration
28 
29You can also call the [RAG Web Browser Actor](https://apify.com/apify/rag-web-browser) directly via its HTTP/SSE interface.
30 
31**Benefits:**
32- ✅ Direct integration without mcp.apify.com
33- ✅ Real-time streaming via Server-Sent Events
34- ✅ Full control over the integration
35- ✅ No additional dependencies
36 
37**Docs:** [Actor Documentation](https://apify.com/apify/rag-web-browser#anthropic-model-context-protocol-mcp-server)
38 
39---
40 
41## 🎯 What does this MCP server do?
42 
43This server is specifically designed to provide fast responses to AI agents and LLMs, allowing them to interact with the web and extract information from web pages.
44It runs locally and communicates with the [RAG Web Browser Actor](https://apify.com/apify/rag-web-browser) in [**Standby mode**](https://docs.apify.com/platform/actors/running/standby),
45sending search queries and receiving extracted web content in response.
46 
47- **Web Search**: Query Google Search, scrape top N URLs, and return cleaned content as Markdown
48- **Single URL Fetching**: Fetch a specific URL and return its content as Markdown
49- **Local MCP Integration**: Standard input/output (stdio) communication with AI clients
50 
51## 🧱 Components
52 
53### Tools
54 
55- name: `search`
56  description: Query Google Search OR fetch a direct URL and return cleaned page contents.
57  arguments:
58  - `query` (string, required): Search keywords or a full URL. Advanced Google operators supported.
59  - `maxResults` (number, optional, default: 1): Max organic results to fetch (ignored when `query` is a URL).
60  - `scrapingTool` (string, optional, default: `raw-http`): One of `browser-playwright` | `raw-http`.
61    - `raw-http`: Fast (no JS execution) – good for static pages.
62    - `browser-playwright`: Handles JS-heavy sites – slower, more robust.
63  - `outputFormats` (array of strings, optional, default: [`markdown`]): One or more of `text`, `markdown`, `html`.
64  - `requestTimeoutSecs` (number, optional, default: 40, min 1 max 300): Total server-side AND client wait budget. A local abort is enforced.
65 
66 
67## 🔄 Migration Guide
68 
69### From Local MCP Server to mcp.apify.com
70 
71**Before (Deprecated local server):**
72```json
73{
74  "mcpServers": {
75    "rag-web-browser": {
76      "command": "npx",
77      "args": ["@apify/mcp-server-rag-web-browser"],
78      "env": {
79        "APIFY_TOKEN": "your-apify-api-token"
80      }
81    }
82  }
83}
84```
85 
86**After (Recommended Apify server):**
87```json
88{
89  "mcpServers": {
90    "apify": {
91      "command": "npx",
92      "args": ["@apify/actors-mcp-server"],
93      "env": {
94        "APIFY_TOKEN": "your-apify-api-token"
95      }
96    }
97  }
98}
99```
100Or use the hosted endpoint: `https://mcp.apify.com` (when your client supports HTTP transport / remote MCP).
101 
102### MCP clients
103- Claude Desktop: https://claude.ai/download
104- Visual Studio Code
105- Apify Tester MCP Client: https://apify.com/jiri.spilka/tester-mcp-client
106 
107## 🛠️ Development
108 
109### Prerequisites
110- Node.js (v18 or higher)
111- Apify API Token (`APIFY_TOKEN`)
112 
113Clone & install:
114```bash
115git clone https://github.com/apify/mcp-server-rag-web-browser.git
116cd mcp-server-rag-web-browser
117npm install
118```
119 
120### Build
121```bash
122npm install
123npm run build
124```
125 
126### Debugging
127 
128Since MCP servers operate over standard input/output (stdio), debugging can be challenging.
129For the best debugging experience, use the [MCP Inspector](https://github.com/modelcontextprotocol/inspector).
130 
131You can launch the MCP Inspector via [`npm`](https://docs.npmjs.com/downloading-and-installing-node-js-and-npm) with this command:
132 
133```bash
134export APIFY_TOKEN=your-apify-api-token
135npx @modelcontextprotocol/inspector node dist/index.js
136```
137Upon launching, the Inspector will display a URL that you can access in your browser to begin debugging.
138 
139# 📖 Learn more
140 
141- [Model Context Protocol](https://modelcontextprotocol.org/)
142- [RAG Web Browser Actor](https://apify.com/apify/rag-web-browser)
143- [What are AI Agents?](https://blog.apify.com/what-are-ai-agents/)
144- [What is MCP and why does it matter?](https://blog.apify.com/what-is-model-context-protocol/)
145- [How to use MCP with Apify Actors](https://blog.apify.com/how-to-use-mcp/)
146- [Tester MCP Client](https://apify.com/jiri.spilka/tester-mcp-client)
147- [Webinar: Building and Monetizing MCP Servers on Apify](https://www.youtube.com/watch?v=w3AH3jIrXXo)
148- [How to build and monetize an AI agent on Apify](https://blog.apify.com/how-to-build-an-ai-agent/)
149- [Build and deploy MCP servers in minutes with a TypeScript template](https://blog.apify.com/build-and-deploy-mcp-servers-typescript/)
150 
151*This repository is maintained for archival purposes only. Please use the recommended alternatives above for active development.*
152