feat(rag): add PowerPoint (.pptx) extraction with VLM image support

docs/guides/chat.mdx

@@ -102,23 +102,24 @@ gaia chat --query "Hello" --show-stats
 ## Document Q&A (RAG)
 
 <Note>
-RAG (Retrieval-Augmented Generation) enables chatting with PDF documents using semantic search and context retrieval.
+RAG (Retrieval-Augmented Generation) enables chatting with PDF and PowerPoint (.pptx) documents using semantic search and context retrieval.
 </Note>
 
 ### CLI with RAG
 
 <Tabs>
   <Tab title="Single Document">
     ```bash
-    # Chat with single document
+    # Chat with a PDF or PowerPoint document
     gaia chat --index manual.pdf
+    gaia chat --index slides.pptx
     ```
   </Tab>
 
   <Tab title="Multiple Documents">
     ```bash
-    # Chat with multiple documents
-    gaia chat --index doc1.pdf doc2.pdf doc3.pdf
+    # Chat with multiple documents (PDF and PPTX supported)
+    gaia chat --index doc1.pdf doc2.pdf slides.pptx
     ```
   </Tab>
 
@@ -131,7 +132,7 @@ RAG (Retrieval-Augmented Generation) enables chatting with PDF documents using s
 
   <Tab title="Watch Folder">
     ```bash
-    # Auto-index every PDF in a folder, and any new ones dropped in later
+    # Auto-index every PDF/PPTX in a folder, and any new ones dropped in later
     gaia chat --watch ./docs
     ```
   </Tab>
@@ -152,7 +153,7 @@ RAG (Retrieval-Augmented Generation) enables chatting with PDF documents using s
 </Tabs>
 
 <Note>
-  **PDF Indexing Requirements:** Processing PDFs with images requires a Vision Language Model (VLM). GAIA uses `Qwen3-VL-4B-Instruct-GGUF` by default for extracting text from images in PDFs.
+  **Document Indexing Requirements:** Processing PDFs and PPTX files with images requires a Vision Language Model (VLM). GAIA uses `Qwen3-VL-4B-Instruct-GGUF` by default for extracting text from images in documents.
 
   To download all models needed for chat (including VLM):
   ```bash

docs/sdk/sdks/rag.mdx

@@ -62,7 +62,7 @@ This happens once when you index documents:
 ```mermaid
 %%{init: {'theme':'base', 'themeVariables': { 'primaryColor':'#ED1C24', 'primaryTextColor':'#fff', 'primaryBorderColor':'#C8171E', 'lineColor':'#F4484D', 'secondaryColor':'#2d2d2d', 'tertiaryColor':'#f5f5f5', 'fontFamily': 'system-ui, -apple-system, sans-serif'}}}%%
 flowchart TD
-    A(["PDF Document"]) --> B(["EXTRACT TEXT"])
+    A(["PDF / PPTX Document"]) --> B(["EXTRACT TEXT"])
     B --> C(["SPLIT INTO CHUNKS"])
     C --> D(["GENERATE EMBEDDINGS"])
     D --> E[("STORE IN FAISS")]

setup.py

@@ -155,6 +155,7 @@
             "numpy>=1.24.0",
             "pymupdf>=1.24.0",
             "pypdf",
+            "python-pptx>=0.6.21",
             "sentence-transformers",
             "safetensors",
             # torch is pinned lower-bound only. The "audio" extra caps
@@ -229,6 +230,7 @@
             "numpy>=1.24.0",
             "pymupdf>=1.24.0",
             "pypdf",
+            "python-pptx>=0.6.21",
             "sentence-transformers",
         ],
         "lint": [