From f8b2e93cc12043222a73ebc89c202fcdf730ac4f Mon Sep 17 00:00:00 2001
From: Vincent <vr@ijpcan.com>
Date: Sat, 22 Aug 2026 10:21:10 +0900
Subject: [PATCH] Add readme file

---
 README.md | 130 ++++++++++++++++++++++++++++++++++++++++++++++++++++++
 1 file changed, 130 insertions(+)
 create mode 100644 README.md

diff --git a/README.md b/README.md
new file mode 100644
index 0000000..3e23543
--- /dev/null
+++ b/README.md
@@ -0,0 +1,130 @@
+# Go through the following steps to use this MCP serer:
+1) Create Python Virtual Env from your root directory
+python -m venv venv
+
+2) Activate Virtual Env
+a) Windows: venv\Scripts\activate.bat
+b) Unix: source venv/bin/activate
+
+3) Installing Dependencies
+pip install -r requirements.txt
+
+4) Start MCP Sever
+python -m mcp_server
+
+
+5) In your project where you want to use this Server, implemt a tool like the following:
+
+@tool("FastMCP Batch Web Content extraction Tool")
+    def fastmcp_batch_web_content_extraction_tool(urls: list[str], query: str) -> str:
+        """
+        Connects directly to the FastMCP HTTP server to scrape and extract text from 
+        a list of multiple URLs simultaneously in parallel.
+        """
+        async def call_fast_mcp():
+            # Initialize client using FastMCP's precise matching transport protocol
+            transport = StreamableHttpTransport("http://localhost:8000/mcp/")
+            
+            async with Client(transport) as client:
+                # call_tool abstracts away deep JSON-RPC structures
+                result = await client.call_tool("batch_web_content_extraction", {"urls": urls, "query": query})
+                return result
+
+        try:
+            # Spin up clean execution loop for this worker thread
+            mcp_response = asyncio.run(call_fast_mcp())
+            
+            # FastMCP Client response objects expose string contents naturally via .content
+            # Ensure we safely extract the raw text string from the MCP content blocks
+            if hasattr(mcp_response, "content") and mcp_response.content:
+                # The text is inside the first content block object
+                raw_content = mcp_response.content[0].text
+            else:
+                # Fallback to string casting if it's already a plain string
+                raw_content = str(mcp_response)
+
+            # Parse JSON and pretty-format the string back to the CrewAI Agent
+            try:
+                parsed_data = json.loads(raw_content)
+                formatted_output = []
+                for item in parsed_data:
+                    formatted_output.append(f"=== SOURCE URL: {item['url']} ===")
+                    formatted_output.append(f"STATUS: {item['status']}")
+                    formatted_output.append(f"CONTENT:\n{item['content']}\n")
+                    formatted_output.append("=" * 40 + "\n")
+
+                logger.info(f"""FastMCP Batch Web Content extraction tool used.\nURLs:\n{"\n".join(urls)}\nQuery: {query}\nResult:\n{"\n".join(formatted_output)}""")    
+                return "\n".join(formatted_output)
+            except json.JSONDecodeError:
+                logger.info(f"""FastMCP Batch Web Content extraction tool used. Error in JSON formatting so raw data returned:\nURLs:\n{"\n".join(urls)}\nQuery: {query}\nResult:\n{raw_content}""")    
+                return raw_content
+                
+        except Exception as e:
+            logger.info(f"""FastMCP Batch Web Content extraction tool used. Error executing the tool and no data returned.""") 
+            return f"Error executing Parallel FastMCP tool over Streamable HTTP: {str(e)}"
+
+
+# Info about this Server
+
+MCP server that fetches multiple URLs in parallel and extracts the content relevant to a query.
+-Clean and deduplicate URLs
+-Clean extracted text
+-Semantic selection to keep only text relevant to the query
+-Limit number of extracted characters per URL and also limit the total characters extracted to avoid returning too much text
+
+Arguments:
+urls:
+List of URLs to process.
+
+query:
+User question or information wanted
+Semantic ranking is performed against this query.
+
+Optional Arguments:
+max_chars_per_url:
+Maximum characters returned for each URL.
+Default set to 6000
+
+max_total_chars:
+Maximum characters returned across all URLs.
+Default set to 30000
+
+top_k_chunks:
+Maximum number of relevant chunks per URL.
+Default set to 6
+
+min_relevance_score:
+Minimum semantic similarity score.
+Default set to 0.25
+
+Returns:
+List of relevant web content results.
+Below an example of info return for one URL:
+
+========================================
+
+=== SOURCE URL: https://www.marketsandmarkets.com/Market-Reports/3d-scanner-market-119952472.html ===
+STATUS: success
+CONTENT:
+[Semantic Retrieval]
+Query: Extract information about companies, products, pricing, key features, target customers, market trends, growth, and competitive information related to 3D scanners.
+Chunks considered: 109
+Chunks selected: 6
+Relevance scores: [0.759, 0.757, 0.749, 0.71, 0.7, 0.696]
+Content characters: 3046
+Truncated: False
+
+--- RELEVANT CONTENT ---
+Chunk 1
+
+Chunk 2
+
+Chunk 3
+
+Chunk 4
+
+Chunk 5
+
+Chunk 6
+
+========================================
\ No newline at end of file