Simplify MCP arch #5

README.md

@@ -63,6 +63,7 @@ tags:
 - Increased max tokens to prevent early stopping
 - Custom EOS handling for medical models
 - Dynamic system prompts based on RAG status
+- One-click agentic toggle to run MedSwin alone (no RAG/web search) for deterministic, offline-safe answers
 
 ## 🚀 Usage
 
@@ -71,6 +72,7 @@ tags:
    - Enable/disable Document RAG
    - Enable/disable Web Search (MCP)
    - Select medical model (MedSwin SFT/KD/TA)
+   - Toggle agentic reasoning on/off for MedSwin-only responses
 3. **Ask Questions**: Type your medical question in any language
 4. **Get Answers**: Receive comprehensive answers based on:
    - Your uploaded documents (if RAG enabled)
@@ -101,6 +103,14 @@ See `requirements.txt` for full dependency list. Key dependencies:
 - **Utilities**: `langdetect`, `gradio`, `spaces`, `soundfile`
 - **TTS**: Optional - `TTS` package (voice features work with MCP fallback if unavailable)
 
+### 🔑 Access Tokens
+
+- **Hugging Face**: Set `HF_TOKEN` so Transformers can download the MedSwin checkpoints. Generate a read-only token at [huggingface.co/settings/tokens](https://huggingface.co/settings/tokens) and export it:
+  ```bash
+  export HF_TOKEN="hf_your_token"
+  ```
+- **Gemini**: `GEMINI_API_KEY` remains required for MCP translation, parsing, transcription, and summarization (see MCP configuration below).
+
 ### 🔌 MCP Configuration
 
 The application uses a bundled Gemini MCP server (agent.py) for translation, document parsing, transcription, and summarization. Configure via environment variables:

agent.py

@@ -215,8 +215,8 @@ async def call_tool(name: str, arguments: dict) -> Sequence[TextContent | ImageC
             
             # Generate content using Gemini API
             try:
-                # Get the model instance
-                gemini_model = gemini_client.models.get(model)
+                # Get the model instance (requires keyword argument)
+                gemini_model = gemini_client.models.get(model=model)
                 
                 # Prepare generation config
                 generation_config = {