Upload Base/Datasets/rag_mcp_sft/source_manifest.json with huggingface_hub

Base/Datasets/rag_mcp_sft/source_manifest.json

@@ -0,0 +1,35 @@
+{
+  "retrieved_on": "2026-04-03",
+  "sources": [
+    {
+      "topic": "mcp",
+      "url": "https://modelcontextprotocol.io/introduction",
+      "title": "What is the Model Context Protocol (MCP)?",
+      "notes": [
+        "MCP is an open standard for connecting AI applications to tools, data sources, and workflows.",
+        "The ecosystem spans hosts like IDEs and assistants plus many local and remote servers.",
+        "The main value is interoperable context and actions across tools without bespoke integrations."
+      ]
+    },
+    {
+      "topic": "mcp",
+      "url": "https://modelcontextprotocol.io/docs/concepts/architecture",
+      "title": "MCP Architecture Overview",
+      "notes": [
+        "MCP uses a client-server architecture with hosts, clients, and servers.",
+        "The data layer is JSON-RPC 2.0 based and includes lifecycle, tools, resources, prompts, and notifications.",
+        "The transport layer covers stdio for local use and streamable HTTP for remote use."
+      ]
+    },
+    {
+      "topic": "rag",
+      "url": "https://www.pinecone.io/learn/retrieval-augmented-generation/",
+      "title": "Retrieval-Augmented Generation (RAG)",
+      "notes": [
+        "RAG addresses stale knowledge, weak domain depth, missing private data, and hallucination risk.",
+        "Core stages are ingestion, retrieval, augmentation, and generation.",
+        "Good RAG systems emphasize chunking, retrieval quality, reranking, grounding, citations, and evaluation."
+      ]
+    }
+  ]
+}