Real-time RAG pipeline pulsing + prominent LangChain branding

- rag-api: new POST /chat/stream SSE endpoint emits live pipeline stages (embed/retrieve/context/llm/answer) and streams LLM tokens
- Knowledge Chat: document chat now streams answers token-by-token and pulses each pipeline stage in real time as it executes
- How-it-works panel: active stage glows/scales, completed stages settle, agent mode also drives the pulse
- LangChain made visible: orchestrated-by-LangChain badge + LC markers on LangChain-native nodes (TextSplitter, Embeddings, as_retriever, ChatOpenAI, Chroma)

ui/src/styles/globals.css

@@ -207,3 +207,13 @@
 @media (prefers-reduced-motion: reduce) {
   .rag-track, .rag-dot, .rag-node { animation: none !important; }
 }
+
+
+/* RAG flow — active/done stage states (real-time per-stage pulsing) */
+@keyframes rag-active-pulse {
+  0%, 100% { box-shadow: 0 0 0 0 var(--rag-glow); transform: scale(1); }
+  50% { box-shadow: 0 0 16px 3px var(--rag-glow); transform: scale(1.08); }
+}
+.rag-node-active { animation: rag-active-pulse 0.8s ease-in-out infinite; border-color: var(--rag-line) !important; z-index: 1; }
+.rag-node-done { border-color: var(--rag-line) !important; }
+@media (prefers-reduced-motion: reduce) { .rag-node-active { animation: none !important; } }