Upload folder using huggingface_hub

nbs/01_clinical_tutor.ipynb

@@ -1,588 +1,575 @@
-{
- "cells": [
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "d9a32ad5-9f07-47f2-97ae-15b0646e355b",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "#| default_exp clinical_tutor"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "0db4b759-310c-4e38-9fdc-2efb94b541dd",
-   "metadata": {},
-   "source": [
-    "# Clinical Tutor\n",
-    "\n",
-    "> Core module for using learning context for context-appropriate tutor responses\n"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "16f93992-88dd-409b-8370-b86302e1ce6a",
-   "metadata": {},
-   "source": [
-    "## Setup"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "6d2403bb-70a1-4744-be0b-d259234c1b62",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "#| hide\n",
-    "from nbdev.showdoc import *"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "477ba22b-55c1-467e-8206-a92f88a598fd",
-   "metadata": {},
-   "outputs": [
-    {
-     "name": "stderr",
-     "output_type": "stream",
-     "text": [
-      "C:\\Users\\deepa\\AppData\\Local\\Programs\\Python\\Python312\\Lib\\site-packages\\tqdm\\auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
-      "  from .autonotebook import tqdm as notebook_tqdm\n"
-     ]
-    }
-   ],
-   "source": [
-    "#| export\n",
-    "from typing import Dict, List, Optional, Any, Tuple\n",
-    "import os\n",
-    "import json\n",
-    "import logging\n",
-    "from pathlib import Path\n",
-    "from datetime import datetime\n",
-    "from dotenv import load_dotenv\n",
-    "import aiohttp\n",
-    "import re\n",
-    "from collections import defaultdict\n",
-    "from wardbuddy.learning_context import LearningContext, setup_logger\n",
-    "\n",
-    "# Load environment variables\n",
-    "load_dotenv()\n",
-    "\n",
-    "logger = setup_logger(__name__)"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "7da445d7-7f51-44d5-b027-e2cf65c79069",
-   "metadata": {},
-   "source": [
-    "## Adaptive Clinical Tutor"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "58d76615-480c-4fe4-9d4a-e67dd892132a",
-   "metadata": {},
-   "source": [
-    "This module implements:\n",
-    "\n",
-    " - Engages in natural case discussions like a clinical supervisor\n",
-    " - Provides context-aware feedback based on student's rotation and preferences\n",
-    " - Analyzes discussions to track learning progress\n",
-    " - Integrates with the student's learning context\n",
-    "\n",
-    "The tutor aims to mimic real-world clinical teaching interactions where students present cases and receive feedback in a natural conversational style.\n"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "da7d2115-b30f-40ba-9566-bcbaf155026d",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "#| export \n",
-    "class OpenRouterException(Exception):\n",
-    "    \"\"\"Custom exception for OpenRouter API errors\"\"\"\n",
-    "    pass"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "75c2cbfc-75e7-45ee-9d86-b931f81a3ad5",
-   "metadata": {},
-   "outputs": [],
-   "source": [
-    "#| export\n",
-    "class ClinicalTutor:\n",
-    "    \"\"\"\n",
-    "    Adaptive clinical teaching module that provides context-aware feedback.\n",
-    "    \n",
-    "    The tutor acts as an experienced clinical supervisor, engaging in natural\n",
-    "    case discussions while tracking student progress and adapting feedback\n",
-    "    based on learning context.\n",
-    "    \n",
-    "    Attributes:\n",
-    "        learning_context (LearningContext): Student's learning context\n",
-    "        model (str): LLM model identifier\n",
-    "        api_url (str): OpenRouter API endpoint\n",
-    "    \"\"\"\n",
-    "    \n",
-    "    def __init__(\n",
-    "        self,\n",
-    "        context_path: Optional[Path] = None,\n",
-    "        model: str = \"anthropic/claude-3.5-sonnet\"\n",
-    "    ):\n",
-    "        \"\"\"\n",
-    "        Initialize clinical tutor.\n",
-    "        \n",
-    "        Args:\n",
-    "            context_path: Optional path for context persistence\n",
-    "            model: Model identifier for OpenRouter\n",
-    "        \"\"\"\n",
-    "        self.api_key: str = os.getenv(\"OPENROUTER_API_KEY\")\n",
-    "        if not self.api_key:\n",
-    "            raise ValueError(\"OpenRouter API key not found\")\n",
-    "        \n",
-    "        self.api_url: str = \"https://openrouter.ai/api/v1/chat/completions\"\n",
-    "        self.model: str = model\n",
-    "        \n",
-    "        self.learning_context = LearningContext(context_path)\n",
-    "        self.context_path = context_path\n",
-    "        \n",
-    "        # Track conversation state\n",
-    "        self.current_case: Dict = {\n",
-    "            \"started\": None,\n",
-    "            \"chief_complaint\": None,\n",
-    "            \"key_findings\": [],\n",
-    "            \"assessment\": None,\n",
-    "            \"plan\": None\n",
-    "        }\n",
-    "        \n",
-    "        logger.info(f\"Clinical tutor initialized with model: {model}\")\n",
-    "    \n",
-    "    async def _get_completion(\n",
-    "        self,\n",
-    "        messages: List[Dict],\n",
-    "        temperature: float = 0.7,\n",
-    "        max_retries: int = 3\n",
-    "    ) -> str:\n",
-    "        \"\"\"\n",
-    "        Get completion from OpenRouter API with retry logic.\n",
-    "        \n",
-    "        Args:\n",
-    "            messages: List of conversation messages\n",
-    "            temperature: Temperature for response generation\n",
-    "            max_retries: Maximum retry attempts\n",
-    "            \n",
-    "        Returns:\n",
-    "            str: Model response\n",
-    "            \n",
-    "        Raises:\n",
-    "            OpenRouterException: If API calls fail after retries\n",
-    "        \"\"\"\n",
-    "        headers = {\n",
-    "            \"Authorization\": f\"Bearer {self.api_key}\",\n",
-    "            \"Content-Type\": \"application/json\",\n",
-    "            \"HTTP-Referer\": \"http://localhost:7860\"\n",
-    "        }\n",
-    "        \n",
-    "        data = {\n",
-    "            \"model\": self.model,\n",
-    "            \"messages\": messages,\n",
-    "            \"temperature\": temperature,\n",
-    "            \"max_tokens\": 2000\n",
-    "        }\n",
-    "        \n",
-    "        for attempt in range(max_retries):\n",
-    "            try:\n",
-    "                async with aiohttp.ClientSession() as session:\n",
-    "                    async with session.post(\n",
-    "                        self.api_url,\n",
-    "                        headers=headers,\n",
-    "                        json=data,\n",
-    "                        timeout=30\n",
-    "                    ) as response:\n",
-    "                        response.raise_for_status()\n",
-    "                        result = await response.json()\n",
-    "                        return result[\"choices\"][0][\"message\"][\"content\"]\n",
-    "                        \n",
-    "            except Exception as e:\n",
-    "                if attempt == max_retries - 1:\n",
-    "                    raise OpenRouterException(f\"API call failed: {str(e)}\")\n",
-    "                logger.warning(f\"Retry {attempt + 1} after error: {str(e)}\")\n",
-    "                # Could add exponential backoff here if needed\n",
-    "    \n",
-    "    def _build_discussion_prompt(self) -> str:\n",
-    "        \"\"\"Build context-aware prompt for case discussion.\"\"\"\n",
-    "        rotation = self.learning_context.current_rotation\n",
-    "        active_preferences = [\n",
-    "            p[\"focus\"] for p in self.learning_context.feedback_preferences \n",
-    "            if p[\"active\"]\n",
-    "        ]\n",
-    "        \n",
-    "        significant_gaps = {\n",
-    "            topic: score for topic, score \n",
-    "            in self.learning_context.knowledge_profile[\"gaps\"].items()\n",
-    "            if score < 0.7  # Only include significant gaps\n",
-    "        }\n",
-    "        \n",
-    "        prompt = f\"\"\"You are an experienced clinical supervisor in {rotation['specialty']}. Act as an engaging and conversational tutor who coaches towards deeper understanding through Socratic dialogue and targeted questions.\n",
-    "\n",
-    "        Key Principles:\n",
-    "        1. Assume I have strong foundational knowledge in medicine, clinical reasoning, and pre-medical sciences\n",
-    "        2. Focus on high-level connections and nuanced clinical decision-making\n",
-    "        3. Use targeted questions to explore my thought process and highlight key learning points\n",
-    "        4. Share relevant clinical pearls and real-world applications\n",
-    "        5. Be conversational and engaging, avoiding lecture-style responses\n",
-    "        \n",
-    "        Current Rotation Focus Areas:\n",
-    "        {', '.join(rotation['key_focus_areas'])}\n",
-    "\n",
-    "        Areas for Deep Dive:\n",
-    "        {', '.join(f'{topic} (confidence: {score:.1f})' for topic, score in significant_gaps.items()) if significant_gaps else 'General clinical reasoning'}\n",
-    "\n",
-    "        Student's Interests:\n",
-    "        {', '.join(active_preferences) if active_preferences else 'Broad clinical discussion'}\n",
-    "\n",
-    "        Engage as a supportive colleague who challenges thinking through conversation. Ask probing questions \n",
-    "        that explore clinical reasoning and highlight important connections. I will ask for clarification \n",
-    "        if concepts need more explanation.\"\"\"\n",
-    "\n",
-    "        return prompt\n",
-    "    \n",
-    "    def _build_analysis_prompt(self, conversation: List[Dict[str, str]]) -> str:\n",
-    "        \"\"\"\n",
-    "        Build prompt for post-discussion analysis.\n",
-    "        \n",
-    "        Args:\n",
-    "            conversation: List of message dictionaries with roles and content\n",
-    "            \n",
-    "        Returns:\n",
-    "            str: Analysis prompt\n",
-    "        \"\"\"\n",
-    "        # Extract case details\n",
-    "        case_content = \"\"\n",
-    "        for msg in conversation:\n",
-    "            if msg[\"role\"] == \"user\":\n",
-    "                case_content += msg[\"content\"] + \"\\n\"\n",
-    "        \n",
-    "        return f\"\"\"Analyze the following case discussion between a medical student and \n",
-    "        clinical supervisor. Focus on the student's demonstrated knowledge, skills, \n",
-    "        and areas for improvement.\n",
-    "\n",
-    "        Case Content:\n",
-    "        {case_content}\n",
-    "\n",
-    "        Please identify:\n",
-    "        1. Key clinical concepts and learning points demonstrated or discussed\n",
-    "        2. Areas where the student showed uncertainty or knowledge gaps\n",
-    "        3. Strengths demonstrated in clinical reasoning and presentation\n",
-    "        4. Specific learning objectives that would help the student's development\n",
-    "\n",
-    "        Frame your response to help with ongoing learning:\n",
-    "        - Start with positive observations\n",
-    "        - Be specific about knowledge gaps\n",
-    "        - Make concrete suggestions for improvement\n",
-    "        - Connect to practical clinical scenarios\"\"\"\n",
-    "    \n",
-    "    async def discuss_case(\n",
-    "        self, \n",
-    "        message: str,\n",
-    "        temperature: float = 0.7\n",
-    "    ) -> str:\n",
-    "        \"\"\"\n",
-    "        Natural case discussion with context-aware responses.\n",
-    "        \n",
-    "        Args:\n",
-    "            message: Student's input message\n",
-    "            temperature: Temperature for response generation\n",
-    "            \n",
-    "        Returns:\n",
-    "            str: Clinical supervisor's response\n",
-    "        \"\"\"\n",
-    "        try:\n",
-    "            # Update case tracking\n",
-    "            if not self.current_case[\"started\"]:\n",
-    "                self.current_case[\"started\"] = datetime.now()\n",
-    "                # Try to identify chief complaint from first message\n",
-    "                cc_match = re.search(r\"(\\d+)\\s*[yY][oO]\\s*[MmFf]\\s*with\\s*([^.]*)\", message)\n",
-    "                if cc_match:\n",
-    "                    self.current_case[\"chief_complaint\"] = cc_match.group(2).strip()\n",
-    "            \n",
-    "            # Build system prompt\n",
-    "            system_prompt = self._build_discussion_prompt()\n",
-    "            \n",
-    "            messages = [{\n",
-    "                \"role\": \"system\",\n",
-    "                \"content\": system_prompt\n",
-    "            }, {\n",
-    "                \"role\": \"user\",\n",
-    "                \"content\": message\n",
-    "            }]\n",
-    "            \n",
-    "            response = await self._get_completion(messages, temperature)\n",
-    "            return response\n",
-    "            \n",
-    "        except Exception as e:\n",
-    "            logger.error(f\"Error in case discussion: {str(e)}\")\n",
-    "            return \"I apologize, but I encountered an error. Please try presenting your case again.\"\n",
-    "    \n",
-    "    async def analyze_discussion(\n",
-    "        self,\n",
-    "        conversation: List[Dict[str, str]]\n",
-    "    ) -> Dict[str, Any]:\n",
-    "        \"\"\"\n",
-    "        Analyze completed case discussion for learning insights.\n",
-    "        \n",
-    "        Args:\n",
-    "            conversation: List of message dictionaries with roles and content\n",
-    "            \n",
-    "        Returns:\n",
-    "            dict: Analysis results containing:\n",
-    "                - learning_points: List of key concepts learned\n",
-    "                - gaps: Dict of identified knowledge gaps\n",
-    "                - strengths: List of demonstrated strengths\n",
-    "                - suggested_objectives: List of recommended learning goals\n",
-    "        \"\"\"\n",
-    "        try:\n",
-    "            # Reset case tracking\n",
-    "            self.current_case = {\n",
-    "                \"started\": None,\n",
-    "                \"chief_complaint\": None,\n",
-    "                \"key_findings\": [],\n",
-    "                \"assessment\": None,\n",
-    "                \"plan\": None\n",
-    "            }\n",
-    "            \n",
-    "            # Get analysis from model\n",
-    "            analysis_prompt = self._build_analysis_prompt(conversation)\n",
-    "            messages = [{\n",
-    "                \"role\": \"system\",\n",
-    "                \"content\": analysis_prompt\n",
-    "            }]\n",
-    "            messages.extend(conversation)\n",
-    "            \n",
-    "            response = await self._get_completion(messages, temperature=0.3)\n",
-    "            \n",
-    "            # Parse insights\n",
-    "            insights = self._parse_analysis(response)\n",
-    "            \n",
-    "            # Update learning context\n",
-    "            self._update_context_from_analysis(insights)\n",
-    "            \n",
-    "            return insights\n",
-    "            \n",
-    "        except Exception as e:\n",
-    "            logger.error(f\"Error in discussion analysis: {str(e)}\")\n",
-    "            return {\n",
-    "                \"learning_points\": [],\n",
-    "                \"gaps\": {},\n",
-    "                \"strengths\": [],\n",
-    "                \"suggested_objectives\": []\n",
-    "            }\n",
-    "    \n",
-    "    def _parse_analysis(self, response: str) -> Dict[str, Any]:\n",
-    "        \"\"\"\n",
-    "        Parse analysis response into structured insights.\n",
-    "        \n",
-    "        Uses pattern matching and basic NLP to extract:\n",
-    "        - Learning points (key concepts discussed)\n",
-    "        - Knowledge gaps with confidence estimates\n",
-    "        - Demonstrated strengths\n",
-    "        - Suggested learning objectives\n",
-    "        \n",
-    "        Args:\n",
-    "            response: Raw analysis response\n",
-    "            \n",
-    "        Returns:\n",
-    "            dict: Structured analysis insights\n",
-    "        \"\"\"\n",
-    "        insights = {\n",
-    "            \"learning_points\": [],\n",
-    "            \"gaps\": {},\n",
-    "            \"strengths\": [],\n",
-    "            \"suggested_objectives\": []\n",
-    "        }\n",
-    "        \n",
-    "        try:\n",
-    "            # Split into sections\n",
-    "            sections = response.lower().split(\"\\n\\n\")\n",
-    "            \n",
-    "            for section in sections:\n",
-    "                if \"learning point\" in section or \"key concept\" in section:\n",
-    "                    # Extract bullet points or numbered items\n",
-    "                    points = re.findall(r\"[-•*]\\s*(.+)$\", section, re.MULTILINE)\n",
-    "                    insights[\"learning_points\"].extend(points)\n",
-    "                    \n",
-    "                elif \"gap\" in section or \"uncertainty\" in section:\n",
-    "                    # Look for topic mentions with confidence indicators\n",
-    "                    gaps = re.findall(\n",
-    "                        r\"(limited|uncertain|unclear|difficulty with)\\s+([^,.]+)\", \n",
-    "                        section\n",
-    "                    )\n",
-    "                    for indicator, topic in gaps:\n",
-    "                        # Estimate confidence based on language\n",
-    "                        confidence = 0.4 if \"limited\" in indicator else 0.6\n",
-    "                        insights[\"gaps\"][topic.strip()] = confidence\n",
-    "                        \n",
-    "                elif \"strength\" in section or \"demonstrated\" in section:\n",
-    "                    # Extract positive mentions\n",
-    "                    strengths = re.findall(r\"[-•*]\\s*(.+)$\", section, re.MULTILINE)\n",
-    "                    insights[\"strengths\"].extend(strengths)\n",
-    "                    \n",
-    "                elif \"objective\" in section or \"suggest\" in section:\n",
-    "                    # Extract recommended objectives\n",
-    "                    objectives = re.findall(r\"[-•*]\\s*(.+)$\", section, re.MULTILINE)\n",
-    "                    insights[\"suggested_objectives\"].extend(objectives)\n",
-    "            \n",
-    "            return insights\n",
-    "            \n",
-    "        except Exception as e:\n",
-    "            logger.error(f\"Error parsing analysis: {str(e)}\")\n",
-    "            return insights\n",
-    "    \n",
-    "    def _update_context_from_analysis(self, insights: Dict[str, Any]) -> None:\n",
-    "        \"\"\"\n",
-    "        Update learning context based on discussion analysis.\n",
-    "        \n",
-    "        Args:\n",
-    "            insights: Dictionary of analysis insights\n",
-    "        \"\"\"\n",
-    "        try:\n",
-    "            # Update knowledge gaps\n",
-    "            for topic, confidence in insights[\"gaps\"].items():\n",
-    "                self.learning_context.update_knowledge_gap(topic, confidence)\n",
-    "            \n",
-    "            # Add strengths\n",
-    "            for strength in insights[\"strengths\"]:\n",
-    "                self.learning_context.add_strength(strength)\n",
-    "            \n",
-    "            # Save context if path provided\n",
-    "            if self.context_path:\n",
-    "                self.learning_context.save_context(self.context_path)\n",
-    "                \n",
-    "        except Exception as e:\n",
-    "            logger.error(f\"Error updating context: {str(e)}\")"
-   ]
-  },
-  {
-   "cell_type": "markdown",
-   "id": "6a2b15f5-6841-43cb-9b57-c0e3f1a0b0c2",
-   "metadata": {},
-   "source": [
-    "## Tests"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "67ee6bde-4ade-448e-a831-86f9f7ae82ea",
-   "metadata": {},
-   "outputs": [
-    {
-     "ename": "RuntimeError",
-     "evalue": "asyncio.run() cannot be called from a running event loop",
-     "output_type": "error",
-     "traceback": [
-      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
-      "\u001b[1;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
-      "Cell \u001b[1;32mIn[7], line 55\u001b[0m\n\u001b[0;32m     53\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;18m__name__\u001b[39m \u001b[38;5;241m==\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m__main__\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n\u001b[0;32m     54\u001b[0m     \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01masyncio\u001b[39;00m\n\u001b[1;32m---> 55\u001b[0m     \u001b[43masyncio\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrun\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtest_clinical_tutor\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n",
-      "File \u001b[1;32m~\\AppData\\Local\\Programs\\Python\\Python312\\Lib\\asyncio\\runners.py:190\u001b[0m, in \u001b[0;36mrun\u001b[1;34m(main, debug, loop_factory)\u001b[0m\n\u001b[0;32m    161\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"Execute the coroutine and return the result.\u001b[39;00m\n\u001b[0;32m    162\u001b[0m \n\u001b[0;32m    163\u001b[0m \u001b[38;5;124;03mThis function runs the passed coroutine, taking care of\u001b[39;00m\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m    186\u001b[0m \u001b[38;5;124;03m    asyncio.run(main())\u001b[39;00m\n\u001b[0;32m    187\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[0;32m    188\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m events\u001b[38;5;241m.\u001b[39m_get_running_loop() \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m    189\u001b[0m     \u001b[38;5;66;03m# fail fast with short traceback\u001b[39;00m\n\u001b[1;32m--> 190\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mRuntimeError\u001b[39;00m(\n\u001b[0;32m    191\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124masyncio.run() cannot be called from a running event loop\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m    193\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m Runner(debug\u001b[38;5;241m=\u001b[39mdebug, loop_factory\u001b[38;5;241m=\u001b[39mloop_factory) \u001b[38;5;28;01mas\u001b[39;00m runner:\n\u001b[0;32m    194\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m runner\u001b[38;5;241m.\u001b[39mrun(main)\n",
-      "\u001b[1;31mRuntimeError\u001b[0m: asyncio.run() cannot be called from a running event loop"
-     ]
-    }
-   ],
-   "source": [
-    "async def test_clinical_tutor():\n",
-    "    \"\"\"Test ClinicalTutor functionality\"\"\"\n",
-    "    if not os.getenv(\"OPENROUTER_API_KEY\"):\n",
-    "        print(\"Skipping tests: No API key\")\n",
-    "        return\n",
-    "        \n",
-    "    tutor = ClinicalTutor()\n",
-    "    \n",
-    "    # Test case discussion\n",
-    "    test_case = \"\"\"\n",
-    "    28yo M with chest pain\n",
-    "    - 2 days duration\n",
-    "    - Sharp, pleuritic\n",
-    "    - No fever or cough\n",
-    "    - Vitals stable\n",
-    "    - Clear exam\n",
-    "    A/P: Likely MSK pain\n",
-    "    \"\"\"\n",
-    "    \n",
-    "    try:\n",
-    "        # Test basic discussion\n",
-    "        response = await tutor.discuss_case(test_case)\n",
-    "        assert isinstance(response, str)\n",
-    "        assert len(response) > 0\n",
-    "        \n",
-    "        # Only assert case tracking if chief complaint was detected\n",
-    "        if tutor.current_case[\"chief_complaint\"]:\n",
-    "            assert \"chest pain\" in tutor.current_case[\"chief_complaint\"].lower()\n",
-    "        \n",
-    "        print(\"Discussion test passed\")\n",
-    "        \n",
-    "        # Test discussion analysis\n",
-    "        conversation = [\n",
-    "            {\"role\": \"user\", \"content\": test_case},\n",
-    "            {\"role\": \"assistant\", \"content\": response}\n",
-    "        ]\n",
-    "        \n",
-    "        analysis = await tutor.analyze_discussion(conversation)\n",
-    "        assert isinstance(analysis, dict)\n",
-    "        assert all(k in analysis for k in [\n",
-    "            'learning_points', 'gaps', 'strengths', 'suggested_objectives'\n",
-    "        ])\n",
-    "        print(\"Analysis test passed\")\n",
-    "        \n",
-    "    except Exception as e:\n",
-    "        print(f\"Test failed: {str(e)}\")\n",
-    "        raise\n",
-    "    \n",
-    "    print(\"All clinical tutor tests passed!\")\n",
-    "\n",
-    "# Run tests\n",
-    "if __name__ == \"__main__\":\n",
-    "    import asyncio\n",
-    "    if not asyncio.get_event_loop().is_running():\n",
-    "        asyncio.run(test_clinical_tutor())"
-   ]
-  },
-  {
-   "cell_type": "code",
-   "execution_count": null,
-   "id": "3f469c37-afe3-4682-9cc4-40326ac21b74",
-   "metadata": {},
-   "outputs": [],
-   "source": []
-  }
- ],
- "metadata": {
-  "kernelspec": {
-   "display_name": "Python 3 (ipykernel)",
-   "language": "python",
-   "name": "python3"
-  },
-  "language_info": {
-   "codemirror_mode": {
-    "name": "ipython",
-    "version": 3
-   },
-   "file_extension": ".py",
-   "mimetype": "text/x-python",
-   "name": "python",
-   "nbconvert_exporter": "python",
-   "pygments_lexer": "ipython3",
-   "version": "3.12.7"
-  }
- },
- "nbformat": 4,
- "nbformat_minor": 5
-}
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "d9a32ad5-9f07-47f2-97ae-15b0646e355b",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#| default_exp clinical_tutor"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "0db4b759-310c-4e38-9fdc-2efb94b541dd",
+   "metadata": {},
+   "source": [
+    "# Clinical Tutor\n",
+    "\n",
+    "> Core module for using learning context for context-appropriate tutor responses\n"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "16f93992-88dd-409b-8370-b86302e1ce6a",
+   "metadata": {},
+   "source": [
+    "## Setup"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "6d2403bb-70a1-4744-be0b-d259234c1b62",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#| hide\n",
+    "from nbdev.showdoc import *"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "477ba22b-55c1-467e-8206-a92f88a598fd",
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "C:\\Users\\deepa\\AppData\\Local\\Programs\\Python\\Python312\\Lib\\site-packages\\tqdm\\auto.py:21: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
+      "  from .autonotebook import tqdm as notebook_tqdm\n"
+     ]
+    }
+   ],
+   "source": [
+    "#| export\n",
+    "from typing import Dict, List, Optional, Any, Tuple\n",
+    "import os\n",
+    "import json\n",
+    "import logging\n",
+    "from pathlib import Path\n",
+    "from datetime import datetime\n",
+    "from dotenv import load_dotenv\n",
+    "import aiohttp\n",
+    "import re\n",
+    "from collections import defaultdict\n",
+    "from wardbuddy.learning_context import LearningContext, setup_logger\n",
+    "\n",
+    "# Load environment variables\n",
+    "load_dotenv()\n",
+    "\n",
+    "logger = setup_logger(__name__)"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "7da445d7-7f51-44d5-b027-e2cf65c79069",
+   "metadata": {},
+   "source": [
+    "## Adaptive Clinical Tutor"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "58d76615-480c-4fe4-9d4a-e67dd892132a",
+   "metadata": {},
+   "source": [
+    "This module implements:\n",
+    "\n",
+    " - Engages in natural case discussions like a clinical supervisor\n",
+    " - Provides context-aware feedback based on student's rotation and preferences\n",
+    " - Analyzes discussions to track learning progress\n",
+    " - Integrates with the student's learning context\n",
+    "\n",
+    "The tutor aims to mimic real-world clinical teaching interactions where students present cases and receive feedback in a natural conversational style.\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "da7d2115-b30f-40ba-9566-bcbaf155026d",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#| export \n",
+    "class OpenRouterException(Exception):\n",
+    "    \"\"\"Custom exception for OpenRouter API errors\"\"\"\n",
+    "    pass"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "75c2cbfc-75e7-45ee-9d86-b931f81a3ad5",
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "#| export\n",
+    "class ClinicalTutor:\n",
+    "    \"\"\"\n",
+    "    Adaptive clinical teaching module that provides context-aware feedback.\n",
+    "    \n",
+    "    The tutor acts as an experienced clinical supervisor, engaging in natural\n",
+    "    case discussions while tracking student progress and adapting feedback\n",
+    "    based on learning context.\n",
+    "    \n",
+    "    Attributes:\n",
+    "        learning_context (LearningContext): Student's learning context\n",
+    "        model (str): LLM model identifier\n",
+    "        api_url (str): OpenRouter API endpoint\n",
+    "    \"\"\"\n",
+    "    \n",
+    "    def __init__(\n",
+    "        self,\n",
+    "        context_path: Optional[Path] = None,\n",
+    "        model: str = \"anthropic/claude-3.5-sonnet\"\n",
+    "    ):\n",
+    "        \"\"\"\n",
+    "        Initialize clinical tutor.\n",
+    "        \n",
+    "        Args:\n",
+    "            context_path: Optional path for context persistence\n",
+    "            model: Model identifier for OpenRouter\n",
+    "        \"\"\"\n",
+    "        self.api_key: str = os.getenv(\"OPENROUTER_API_KEY\")\n",
+    "        if not self.api_key:\n",
+    "            raise ValueError(\"OpenRouter API key not found\")\n",
+    "        \n",
+    "        self.api_url: str = \"https://openrouter.ai/api/v1/chat/completions\"\n",
+    "        self.model: str = model\n",
+    "        \n",
+    "        self.learning_context = LearningContext(context_path)\n",
+    "        self.context_path = context_path\n",
+    "        \n",
+    "        # Track conversation state\n",
+    "        self.current_case: Dict = {\n",
+    "            \"started\": None,\n",
+    "            \"chief_complaint\": None,\n",
+    "            \"key_findings\": [],\n",
+    "            \"assessment\": None,\n",
+    "            \"plan\": None\n",
+    "        }\n",
+    "        \n",
+    "        logger.info(f\"Clinical tutor initialized with model: {model}\")\n",
+    "    \n",
+    "    async def _get_completion(\n",
+    "        self,\n",
+    "        messages: List[Dict],\n",
+    "        temperature: float = 0.7,\n",
+    "        max_retries: int = 3\n",
+    "    ) -> str:\n",
+    "        \"\"\"\n",
+    "        Get completion from OpenRouter API with retry logic.\n",
+    "        \n",
+    "        Args:\n",
+    "            messages: List of conversation messages\n",
+    "            temperature: Temperature for response generation\n",
+    "            max_retries: Maximum retry attempts\n",
+    "            \n",
+    "        Returns:\n",
+    "            str: Model response\n",
+    "            \n",
+    "        Raises:\n",
+    "            OpenRouterException: If API calls fail after retries\n",
+    "        \"\"\"\n",
+    "        headers = {\n",
+    "            \"Authorization\": f\"Bearer {self.api_key}\",\n",
+    "            \"Content-Type\": \"application/json\",\n",
+    "            \"HTTP-Referer\": \"http://localhost:7860\"\n",
+    "        }\n",
+    "        \n",
+    "        data = {\n",
+    "            \"model\": self.model,\n",
+    "            \"messages\": messages,\n",
+    "            \"temperature\": temperature,\n",
+    "            \"max_tokens\": 2000\n",
+    "        }\n",
+    "        \n",
+    "        for attempt in range(max_retries):\n",
+    "            try:\n",
+    "                async with aiohttp.ClientSession() as session:\n",
+    "                    async with session.post(\n",
+    "                        self.api_url,\n",
+    "                        headers=headers,\n",
+    "                        json=data,\n",
+    "                        timeout=30\n",
+    "                    ) as response:\n",
+    "                        response.raise_for_status()\n",
+    "                        result = await response.json()\n",
+    "                        return result[\"choices\"][0][\"message\"][\"content\"]\n",
+    "                        \n",
+    "            except Exception as e:\n",
+    "                if attempt == max_retries - 1:\n",
+    "                    raise OpenRouterException(f\"API call failed: {str(e)}\")\n",
+    "                logger.warning(f\"Retry {attempt + 1} after error: {str(e)}\")\n",
+    "                # Could add exponential backoff here if needed\n",
+    "    \n",
+    "    def _build_discussion_prompt(self) -> str:\n",
+    "        \"\"\"Build context-aware prompt for case discussion.\"\"\"\n",
+    "        rotation = self.learning_context.current_rotation\n",
+    "        active_preferences = [\n",
+    "            p[\"focus\"] for p in self.learning_context.feedback_preferences \n",
+    "            if p[\"active\"]\n",
+    "        ]\n",
+    "        \n",
+    "        significant_gaps = {\n",
+    "            topic: score for topic, score \n",
+    "            in self.learning_context.knowledge_profile[\"gaps\"].items()\n",
+    "            if score < 0.7  # Only include significant gaps\n",
+    "        }\n",
+    "        \n",
+    "        prompt = f\"\"\"You are an experienced clinical supervisor in {rotation['specialty']}. Act as an engaging and conversational tutor who coaches towards deeper understanding through Socratic dialogue and targeted questions.\n",
+    "\n",
+    "        Key Principles:\n",
+    "        1. Assume I have strong foundational knowledge in medicine, clinical reasoning, and pre-medical sciences\n",
+    "        2. Focus on high-level connections and nuanced clinical decision-making\n",
+    "        3. Use targeted questions to explore my thought process and highlight key learning points\n",
+    "        4. Share relevant clinical pearls and real-world applications\n",
+    "        5. Be conversational and engaging, avoiding lecture-style responses\n",
+    "        \n",
+    "        Current Rotation Focus Areas:\n",
+    "        {', '.join(rotation['key_focus_areas'])}\n",
+    "\n",
+    "        Areas for Deep Dive:\n",
+    "        {', '.join(f'{topic} (confidence: {score:.1f})' for topic, score in significant_gaps.items()) if significant_gaps else 'General clinical reasoning'}\n",
+    "\n",
+    "        Student's Interests:\n",
+    "        {', '.join(active_preferences) if active_preferences else 'Broad clinical discussion'}\n",
+    "\n",
+    "        Ask probing questions that explore clinical reasoning and highlight important connections. I will ask for clarification \n",
+    "        if concepts need more explanation.\"\"\"\n",
+    "\n",
+    "        return prompt\n",
+    "    \n",
+    "    def _build_analysis_prompt(self, conversation: List[Dict[str, str]]) -> str:\n",
+    "        \"\"\"\n",
+    "        Build prompt for post-discussion analysis.\n",
+    "        \n",
+    "        Args:\n",
+    "            conversation: List of message dictionaries with roles and content\n",
+    "            \n",
+    "        Returns:\n",
+    "            str: Analysis prompt\n",
+    "        \"\"\"\n",
+    "        # Extract case details\n",
+    "        case_content = \"\"\n",
+    "        for msg in conversation:\n",
+    "            if msg[\"role\"] == \"user\":\n",
+    "                case_content += msg[\"content\"] + \"\\n\"\n",
+    "        \n",
+    "        return f\"\"\"Analyze the following case discussion between a medical student and \n",
+    "        clinical supervisor. Focus on the student's demonstrated knowledge, skills, \n",
+    "        and areas for improvement.\n",
+    "\n",
+    "        Case Content:\n",
+    "        {case_content}\n",
+    "\n",
+    "        Please identify:\n",
+    "        1. Key clinical concepts and learning points demonstrated or discussed\n",
+    "        2. Areas where the student showed uncertainty or knowledge gaps\n",
+    "        3. Strengths demonstrated in clinical reasoning and presentation\n",
+    "        4. Specific learning objectives that would help the student's development\n",
+    "\n",
+    "        Frame your response to help with ongoing learning:\n",
+    "        - Start with positive observations\n",
+    "        - Be specific about knowledge gaps\n",
+    "        - Make concrete suggestions for improvement\n",
+    "        - Connect to practical clinical scenarios\"\"\"\n",
+    "    \n",
+    "    async def discuss_case(\n",
+    "        self, \n",
+    "        message: str,\n",
+    "        temperature: float = 0.7\n",
+    "    ) -> str:\n",
+    "        \"\"\"\n",
+    "        Natural case discussion with context-aware responses.\n",
+    "        \n",
+    "        Args:\n",
+    "            message: Student's input message\n",
+    "            temperature: Temperature for response generation\n",
+    "            \n",
+    "        Returns:\n",
+    "            str: Clinical supervisor's response\n",
+    "        \"\"\"\n",
+    "        try:\n",
+    "            # Update case tracking\n",
+    "            if not self.current_case[\"started\"]:\n",
+    "                self.current_case[\"started\"] = datetime.now()\n",
+    "                # Try to identify chief complaint from first message\n",
+    "                cc_match = re.search(r\"(\\d+)\\s*[yY][oO]\\s*[MmFf]\\s*with\\s*([^.]*)\", message)\n",
+    "                if cc_match:\n",
+    "                    self.current_case[\"chief_complaint\"] = cc_match.group(2).strip()\n",
+    "            \n",
+    "            # Build system prompt\n",
+    "            system_prompt = self._build_discussion_prompt()\n",
+    "            \n",
+    "            messages = [{\n",
+    "                \"role\": \"system\",\n",
+    "                \"content\": system_prompt\n",
+    "            }, {\n",
+    "                \"role\": \"user\",\n",
+    "                \"content\": message\n",
+    "            }]\n",
+    "            \n",
+    "            response = await self._get_completion(messages, temperature)\n",
+    "            return response\n",
+    "            \n",
+    "        except Exception as e:\n",
+    "            logger.error(f\"Error in case discussion: {str(e)}\")\n",
+    "            return \"I apologize, but I encountered an error. Please try presenting your case again.\"\n",
+    "    \n",
+    "    async def analyze_discussion(\n",
+    "        self,\n",
+    "        conversation: List[Dict[str, str]]\n",
+    "    ) -> Dict[str, Any]:\n",
+    "        \"\"\"\n",
+    "        Analyze completed case discussion for learning insights.\n",
+    "        \n",
+    "        Args:\n",
+    "            conversation: List of message dictionaries with roles and content\n",
+    "            \n",
+    "        Returns:\n",
+    "            dict: Analysis results containing:\n",
+    "                - learning_points: List of key concepts learned\n",
+    "                - gaps: Dict of identified knowledge gaps\n",
+    "                - strengths: List of demonstrated strengths\n",
+    "                - suggested_objectives: List of recommended learning goals\n",
+    "        \"\"\"\n",
+    "        try:\n",
+    "            # Reset case tracking\n",
+    "            self.current_case = {\n",
+    "                \"started\": None,\n",
+    "                \"chief_complaint\": None,\n",
+    "                \"key_findings\": [],\n",
+    "                \"assessment\": None,\n",
+    "                \"plan\": None\n",
+    "            }\n",
+    "            \n",
+    "            # Get analysis from model\n",
+    "            analysis_prompt = self._build_analysis_prompt(conversation)\n",
+    "            messages = [{\n",
+    "                \"role\": \"system\",\n",
+    "                \"content\": analysis_prompt\n",
+    "            }]\n",
+    "            messages.extend(conversation)\n",
+    "            \n",
+    "            response = await self._get_completion(messages, temperature=0.3)\n",
+    "            \n",
+    "            # Parse insights\n",
+    "            insights = self._parse_analysis(response)\n",
+    "            \n",
+    "            # Update learning context\n",
+    "            self._update_context_from_analysis(insights)\n",
+    "            \n",
+    "            return insights\n",
+    "            \n",
+    "        except Exception as e:\n",
+    "            logger.error(f\"Error in discussion analysis: {str(e)}\")\n",
+    "            return {\n",
+    "                \"learning_points\": [],\n",
+    "                \"gaps\": {},\n",
+    "                \"strengths\": [],\n",
+    "                \"suggested_objectives\": []\n",
+    "            }\n",
+    "    \n",
+    "    def _parse_analysis(self, response: str) -> Dict[str, Any]:\n",
+    "        \"\"\"\n",
+    "        Parse analysis response into structured insights.\n",
+    "        \n",
+    "        Uses pattern matching and basic NLP to extract:\n",
+    "        - Learning points (key concepts discussed)\n",
+    "        - Knowledge gaps with confidence estimates\n",
+    "        - Demonstrated strengths\n",
+    "        - Suggested learning objectives\n",
+    "        \n",
+    "        Args:\n",
+    "            response: Raw analysis response\n",
+    "            \n",
+    "        Returns:\n",
+    "            dict: Structured analysis insights\n",
+    "        \"\"\"\n",
+    "        insights = {\n",
+    "            \"learning_points\": [],\n",
+    "            \"gaps\": {},\n",
+    "            \"strengths\": [],\n",
+    "            \"suggested_objectives\": []\n",
+    "        }\n",
+    "        \n",
+    "        try:\n",
+    "            # Split into sections\n",
+    "            sections = response.lower().split(\"\\n\\n\")\n",
+    "            \n",
+    "            for section in sections:\n",
+    "                if \"learning point\" in section or \"key concept\" in section:\n",
+    "                    # Extract bullet points or numbered items\n",
+    "                    points = re.findall(r\"[-•*]\\s*(.+)$\", section, re.MULTILINE)\n",
+    "                    insights[\"learning_points\"].extend(points)\n",
+    "                    \n",
+    "                elif \"gap\" in section or \"uncertainty\" in section:\n",
+    "                    # Look for topic mentions with confidence indicators\n",
+    "                    gaps = re.findall(\n",
+    "                        r\"(limited|uncertain|unclear|difficulty with)\\s+([^,.]+)\", \n",
+    "                        section\n",
+    "                    )\n",
+    "                    for indicator, topic in gaps:\n",
+    "                        # Estimate confidence based on language\n",
+    "                        confidence = 0.4 if \"limited\" in indicator else 0.6\n",
+    "                        insights[\"gaps\"][topic.strip()] = confidence\n",
+    "                        \n",
+    "                elif \"strength\" in section or \"demonstrated\" in section:\n",
+    "                    # Extract positive mentions\n",
+    "                    strengths = re.findall(r\"[-•*]\\s*(.+)$\", section, re.MULTILINE)\n",
+    "                    insights[\"strengths\"].extend(strengths)\n",
+    "                    \n",
+    "                elif \"objective\" in section or \"suggest\" in section:\n",
+    "                    # Extract recommended objectives\n",
+    "                    objectives = re.findall(r\"[-•*]\\s*(.+)$\", section, re.MULTILINE)\n",
+    "                    insights[\"suggested_objectives\"].extend(objectives)\n",
+    "            \n",
+    "            return insights\n",
+    "            \n",
+    "        except Exception as e:\n",
+    "            logger.error(f\"Error parsing analysis: {str(e)}\")\n",
+    "            return insights\n",
+    "    \n",
+    "    def _update_context_from_analysis(self, insights: Dict[str, Any]) -> None:\n",
+    "        \"\"\"\n",
+    "        Update learning context based on discussion analysis.\n",
+    "        \n",
+    "        Args:\n",
+    "            insights: Dictionary of analysis insights\n",
+    "        \"\"\"\n",
+    "        try:\n",
+    "            # Update knowledge gaps\n",
+    "            for topic, confidence in insights[\"gaps\"].items():\n",
+    "                self.learning_context.update_knowledge_gap(topic, confidence)\n",
+    "            \n",
+    "            # Add strengths\n",
+    "            for strength in insights[\"strengths\"]:\n",
+    "                self.learning_context.add_strength(strength)\n",
+    "            \n",
+    "            # Save context if path provided\n",
+    "            if self.context_path:\n",
+    "                self.learning_context.save_context(self.context_path)\n",
+    "                \n",
+    "        except Exception as e:\n",
+    "            logger.error(f\"Error updating context: {str(e)}\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "id": "6a2b15f5-6841-43cb-9b57-c0e3f1a0b0c2",
+   "metadata": {},
+   "source": [
+    "## Tests"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "67ee6bde-4ade-448e-a831-86f9f7ae82ea",
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "RuntimeError",
+     "evalue": "asyncio.run() cannot be called from a running event loop",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[1;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[1;31mRuntimeError\u001b[0m                              Traceback (most recent call last)",
+      "Cell \u001b[1;32mIn[7], line 55\u001b[0m\n\u001b[0;32m     53\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;18m__name__\u001b[39m \u001b[38;5;241m==\u001b[39m \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124m__main__\u001b[39m\u001b[38;5;124m\"\u001b[39m:\n\u001b[0;32m     54\u001b[0m     \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01masyncio\u001b[39;00m\n\u001b[1;32m---> 55\u001b[0m     \u001b[43masyncio\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mrun\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtest_clinical_tutor\u001b[49m\u001b[43m(\u001b[49m\u001b[43m)\u001b[49m\u001b[43m)\u001b[49m\n",
+      "File \u001b[1;32m~\\AppData\\Local\\Programs\\Python\\Python312\\Lib\\asyncio\\runners.py:190\u001b[0m, in \u001b[0;36mrun\u001b[1;34m(main, debug, loop_factory)\u001b[0m\n\u001b[0;32m    161\u001b[0m \u001b[38;5;250m\u001b[39m\u001b[38;5;124;03m\"\"\"Execute the coroutine and return the result.\u001b[39;00m\n\u001b[0;32m    162\u001b[0m \n\u001b[0;32m    163\u001b[0m \u001b[38;5;124;03mThis function runs the passed coroutine, taking care of\u001b[39;00m\n\u001b[1;32m   (...)\u001b[0m\n\u001b[0;32m    186\u001b[0m \u001b[38;5;124;03m    asyncio.run(main())\u001b[39;00m\n\u001b[0;32m    187\u001b[0m \u001b[38;5;124;03m\"\"\"\u001b[39;00m\n\u001b[0;32m    188\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m events\u001b[38;5;241m.\u001b[39m_get_running_loop() \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n\u001b[0;32m    189\u001b[0m     \u001b[38;5;66;03m# fail fast with short traceback\u001b[39;00m\n\u001b[1;32m--> 190\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mRuntimeError\u001b[39;00m(\n\u001b[0;32m    191\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124masyncio.run() cannot be called from a running event loop\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m    193\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m Runner(debug\u001b[38;5;241m=\u001b[39mdebug, loop_factory\u001b[38;5;241m=\u001b[39mloop_factory) \u001b[38;5;28;01mas\u001b[39;00m runner:\n\u001b[0;32m    194\u001b[0m     \u001b[38;5;28;01mreturn\u001b[39;00m runner\u001b[38;5;241m.\u001b[39mrun(main)\n",
+      "\u001b[1;31mRuntimeError\u001b[0m: asyncio.run() cannot be called from a running event loop"
+     ]
+    }
+   ],
+   "source": [
+    "async def test_clinical_tutor():\n",
+    "    \"\"\"Test ClinicalTutor functionality\"\"\"\n",
+    "    if not os.getenv(\"OPENROUTER_API_KEY\"):\n",
+    "        print(\"Skipping tests: No API key\")\n",
+    "        return\n",
+    "        \n",
+    "    tutor = ClinicalTutor()\n",
+    "    \n",
+    "    # Test case discussion\n",
+    "    test_case = \"\"\"\n",
+    "    28yo M with chest pain\n",
+    "    - 2 days duration\n",
+    "    - Sharp, pleuritic\n",
+    "    - No fever or cough\n",
+    "    - Vitals stable\n",
+    "    - Clear exam\n",
+    "    A/P: Likely MSK pain\n",
+    "    \"\"\"\n",
+    "    \n",
+    "    try:\n",
+    "        # Test basic discussion\n",
+    "        response = await tutor.discuss_case(test_case)\n",
+    "        assert isinstance(response, str)\n",
+    "        assert len(response) > 0\n",
+    "        \n",
+    "        # Only assert case tracking if chief complaint was detected\n",
+    "        if tutor.current_case[\"chief_complaint\"]:\n",
+    "            assert \"chest pain\" in tutor.current_case[\"chief_complaint\"].lower()\n",
+    "        \n",
+    "        print(\"Discussion test passed\")\n",
+    "        \n",
+    "        # Test discussion analysis\n",
+    "        conversation = [\n",
+    "            {\"role\": \"user\", \"content\": test_case},\n",
+    "            {\"role\": \"assistant\", \"content\": response}\n",
+    "        ]\n",
+    "        \n",
+    "        analysis = await tutor.analyze_discussion(conversation)\n",
+    "        assert isinstance(analysis, dict)\n",
+    "        assert all(k in analysis for k in [\n",
+    "            'learning_points', 'gaps', 'strengths', 'suggested_objectives'\n",
+    "        ])\n",
+    "        print(\"Analysis test passed\")\n",
+    "        \n",
+    "    except Exception as e:\n",
+    "        print(f\"Test failed: {str(e)}\")\n",
+    "        raise\n",
+    "    \n",
+    "    print(\"All clinical tutor tests passed!\")\n",
+    "\n",
+    "# Run tests\n",
+    "if __name__ == \"__main__\":\n",
+    "    import asyncio\n",
+    "    if not asyncio.get_event_loop().is_running():\n",
+    "        asyncio.run(test_clinical_tutor())"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "id": "3f469c37-afe3-4682-9cc4-40326ac21b74",
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "python3",
+   "language": "python",
+   "name": "python3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}

nbs/02_learning_interface.ipynb

@@ -90,49 +90,111 @@
     "def create_dashboard_css() -> str:\n",
     "    \"\"\"Create custom CSS for dashboard styling\"\"\"\n",
     "    return \"\"\"\n",
+    "    /* Global styles */\n",
+    "    .gradio-container {\n",
+    "        background-color: #0f172a !important;  /* slate-900 */\n",
+    "    }\n",
+    "    \n",
+    "    /* Card styling */\n",
     "    .dashboard-card {\n",
-    "        border: 1px solid #e2e8f0;\n",
-    "        border-radius: 8px;\n",
-    "        padding: 16px;\n",
-    "        margin: 8px 0;\n",
-    "        background: white;\n",
+    "        background-color: #1e293b !important;  /* slate-800 */\n",
+    "        border: 1px solid #334155 !important;  /* slate-700 */\n",
+    "        border-radius: 0.5rem !important;\n",
+    "        padding: 1rem !important;\n",
+    "        margin: 0.5rem 0 !important;\n",
+    "        color: #f1f5f9 !important;  /* slate-100 */\n",
     "    }\n",
     "    \n",
-    "    .status-active {\n",
-    "        color: #48bb78;\n",
-    "        font-weight: 500;\n",
+    "    /* Chat container */\n",
+    "    .chatbot {\n",
+    "        background-color: #1e293b !important;  /* slate-800 */\n",
+    "        border-color: #334155 !important;  /* slate-700 */\n",
     "    }\n",
     "    \n",
-    "    .status-completed {\n",
-    "        color: #718096;\n",
+    "    /* Message bubbles */\n",
+    "    .chatbot .message.user {\n",
+    "        background-color: #334155 !important;  /* slate-700 */\n",
+    "        border: 1px solid #475569 !important;  /* slate-600 */\n",
+    "        color: #f1f5f9 !important;  /* slate-100 */\n",
     "    }\n",
     "    \n",
-    "    .dashboard-header {\n",
-    "        font-size: 1.25rem;\n",
-    "        font-weight: 600;\n",
-    "        margin-bottom: 1rem;\n",
+    "    .chatbot .message.bot {\n",
+    "        background-color: #1e40af !important;  /* blue-800 */\n",
+    "        border: 1px solid #1e3a8a !important;  /* blue-900 */\n",
+    "        color: #f1f5f9 !important;  /* slate-100 */\n",
+    "    }\n",
+    "    \n",
+    "    /* Input fields */\n",
+    "    textarea, input[type=\"text\"] {\n",
+    "        background-color: #334155 !important;  /* slate-700 */\n",
+    "        color: #f1f5f9 !important;  /* slate-100 */\n",
+    "        border: 1px solid #475569 !important;  /* slate-600 */\n",
+    "    }\n",
+    "    \n",
+    "    textarea:focus, input[type=\"text\"]:focus {\n",
+    "        border-color: #3b82f6 !important;  /* blue-500 */\n",
+    "        box-shadow: 0 0 0 2px rgba(59, 130, 246, 0.2) !important;\n",
+    "    }\n",
+    "    \n",
+    "    /* Buttons */\n",
+    "    button.primary {\n",
+    "        background-color: #2563eb !important;  /* blue-600 */\n",
+    "        color: white !important;\n",
+    "    }\n",
+    "    \n",
+    "    button.primary:hover {\n",
+    "        background-color: #3b82f6 !important;  /* blue-500 */\n",
+    "    }\n",
+    "    \n",
+    "    button.secondary {\n",
+    "        background-color: #475569 !important;  /* slate-600 */\n",
+    "        color: white !important;\n",
+    "    }\n",
+    "    \n",
+    "    button.secondary:hover {\n",
+    "        background-color: #64748b !important;  /* slate-500 */\n",
+    "    }\n",
+    "    \n",
+    "    /* Tabs */\n",
+    "    .tab-nav {\n",
+    "        background-color: #1e293b !important;  /* slate-800 */\n",
+    "        border-bottom: 1px solid #334155 !important;  /* slate-700 */\n",
+    "    }\n",
+    "    \n",
+    "    .tab-nav button {\n",
+    "        color: #f1f5f9 !important;  /* slate-100 */\n",
     "    }\n",
     "    \n",
-    "    .summary-modal {\n",
-    "        max-width: 800px !important;\n",
+    "    .tab-nav button.selected {\n",
+    "        border-bottom-color: #3b82f6 !important;  /* blue-500 */\n",
     "    }\n",
     "    \n",
-    "    .feedback-tag {\n",
-    "        display: inline-block;\n",
-    "        padding: 4px 8px;\n",
-    "        border-radius: 4px;\n",
-    "        margin: 4px;\n",
-    "        font-size: 0.875rem;\n",
+    "    /* Status indicators */\n",
+    "    .status-active {\n",
+    "        color: #22c55e !important;  /* green-500 */\n",
+    "        font-weight: 500 !important;\n",
+    "    }\n",
+    "    \n",
+    "    .status-completed {\n",
+    "        color: #94a3b8 !important;  /* slate-400 */\n",
+    "    }\n",
+    "    \n",
+    "    /* Headers */\n",
+    "    .dashboard-header {\n",
+    "        color: #f1f5f9 !important;  /* slate-100 */\n",
+    "        font-size: 1.5rem !important;\n",
+    "        font-weight: 600 !important;\n",
+    "        margin-bottom: 1rem !important;\n",
     "    }\n",
     "    \n",
-    "    .feedback-active {\n",
-    "        background: #ebf8ff;\n",
-    "        color: #2b6cb0;\n",
+    "    /* Tables */\n",
+    "    table {\n",
+    "        background-color: #1e293b !important;  /* slate-800 */\n",
+    "        color: #f1f5f9 !important;  /* slate-100 */\n",
     "    }\n",
     "    \n",
-    "    .feedback-inactive {\n",
-    "        background: #f7fafc;\n",
-    "        color: #718096;\n",
+    "    th, td {\n",
+    "        border-color: #334155 !important;  /* slate-700 */\n",
     "    }\n",
     "    \"\"\""
    ]

wardbuddy/clinical_tutor.py

@@ -129,52 +129,40 @@ class ClinicalTutor:
                 # Could add exponential backoff here if needed
     
     def _build_discussion_prompt(self) -> str:
-        """
-        Build context-aware prompt for case discussion.
-        
-        Incorporates:
-        - Current rotation details
-        - Active feedback preferences
-        - Recent learning points
-        - Knowledge gaps needing attention
-        
-        Returns:
-            str: Contextualized system prompt
-        """
+        """Build context-aware prompt for case discussion."""
         rotation = self.learning_context.current_rotation
         active_preferences = [
             p["focus"] for p in self.learning_context.feedback_preferences 
             if p["active"]
         ]
         
-        # Get relevant knowledge gaps
         significant_gaps = {
             topic: score for topic, score 
             in self.learning_context.knowledge_profile["gaps"].items()
             if score < 0.7  # Only include significant gaps
         }
         
-        prompt = f"""You are an experienced clinical supervisor in {rotation['specialty']} 
-        providing teaching and feedback. You aim to:
-
-        1. Help students develop strong clinical reasoning
-        2. Connect theory to practical applications
-        3. Build diagnostic confidence
-        4. Improve presentation skills
+        prompt = f"""You are an experienced clinical supervisor in {rotation['specialty']}. Act as an engaging and conversational tutor who coaches towards deeper understanding through Socratic dialogue and targeted questions.
 
+        Key Principles:
+        1. Assume I have strong foundational knowledge in medicine, clinical reasoning, and pre-medical sciences
+        2. Focus on high-level connections and nuanced clinical decision-making
+        3. Use targeted questions to explore my thought process and highlight key learning points
+        4. Share relevant clinical pearls and real-world applications
+        5. Be conversational and engaging, avoiding lecture-style responses
+        
         Current Rotation Focus Areas:
         {', '.join(rotation['key_focus_areas'])}
 
-        Areas Needing Attention:
-        {', '.join(f'{topic} (confidence: {score:.1f})' for topic, score in significant_gaps.items()) if significant_gaps else 'No specific gaps identified'}
+        Areas for Deep Dive:
+        {', '.join(f'{topic} (confidence: {score:.1f})' for topic, score in significant_gaps.items()) if significant_gaps else 'General clinical reasoning'}
 
-        Student's Requested Focus:
-        {', '.join(active_preferences) if active_preferences else 'General clinical feedback'}
+        Student's Interests:
+        {', '.join(active_preferences) if active_preferences else 'Broad clinical discussion'}
+
+        Ask probing questions that explore clinical reasoning and highlight important connections. I will ask for clarification 
+        if concepts need more explanation."""
 
-        Engage naturally as a supportive but challenging supervisor would during case 
-        presentations. Ask probing questions when appropriate, share relevant clinical 
-        pearls, and help the student build their clinical reasoning skills."""
-        
         return prompt
     
     def _build_analysis_prompt(self, conversation: List[Dict[str, str]]) -> str:

wardbuddy/learning_interface.py

@@ -21,49 +21,111 @@ logger = setup_logger(__name__)
 def create_dashboard_css() -> str:
     """Create custom CSS for dashboard styling"""
     return """
+    /* Global styles */
+    .gradio-container {
+        background-color: #0f172a !important;  /* slate-900 */
+    }
+    
+    /* Card styling */
     .dashboard-card {
-        border: 1px solid #e2e8f0;
-        border-radius: 8px;
-        padding: 16px;
-        margin: 8px 0;
-        background: white;
+        background-color: #1e293b !important;  /* slate-800 */
+        border: 1px solid #334155 !important;  /* slate-700 */
+        border-radius: 0.5rem !important;
+        padding: 1rem !important;
+        margin: 0.5rem 0 !important;
+        color: #f1f5f9 !important;  /* slate-100 */
     }
     
-    .status-active {
-        color: #48bb78;
-        font-weight: 500;
+    /* Chat container */
+    .chatbot {
+        background-color: #1e293b !important;  /* slate-800 */
+        border-color: #334155 !important;  /* slate-700 */
     }
     
-    .status-completed {
-        color: #718096;
+    /* Message bubbles */
+    .chatbot .message.user {
+        background-color: #334155 !important;  /* slate-700 */
+        border: 1px solid #475569 !important;  /* slate-600 */
+        color: #f1f5f9 !important;  /* slate-100 */
     }
     
-    .dashboard-header {
-        font-size: 1.25rem;
-        font-weight: 600;
-        margin-bottom: 1rem;
+    .chatbot .message.bot {
+        background-color: #1e40af !important;  /* blue-800 */
+        border: 1px solid #1e3a8a !important;  /* blue-900 */
+        color: #f1f5f9 !important;  /* slate-100 */
+    }
+    
+    /* Input fields */
+    textarea, input[type="text"] {
+        background-color: #334155 !important;  /* slate-700 */
+        color: #f1f5f9 !important;  /* slate-100 */
+        border: 1px solid #475569 !important;  /* slate-600 */
+    }
+    
+    textarea:focus, input[type="text"]:focus {
+        border-color: #3b82f6 !important;  /* blue-500 */
+        box-shadow: 0 0 0 2px rgba(59, 130, 246, 0.2) !important;
+    }
+    
+    /* Buttons */
+    button.primary {
+        background-color: #2563eb !important;  /* blue-600 */
+        color: white !important;
+    }
+    
+    button.primary:hover {
+        background-color: #3b82f6 !important;  /* blue-500 */
+    }
+    
+    button.secondary {
+        background-color: #475569 !important;  /* slate-600 */
+        color: white !important;
+    }
+    
+    button.secondary:hover {
+        background-color: #64748b !important;  /* slate-500 */
+    }
+    
+    /* Tabs */
+    .tab-nav {
+        background-color: #1e293b !important;  /* slate-800 */
+        border-bottom: 1px solid #334155 !important;  /* slate-700 */
+    }
+    
+    .tab-nav button {
+        color: #f1f5f9 !important;  /* slate-100 */
     }
     
-    .summary-modal {
-        max-width: 800px !important;
+    .tab-nav button.selected {
+        border-bottom-color: #3b82f6 !important;  /* blue-500 */
     }
     
-    .feedback-tag {
-        display: inline-block;
-        padding: 4px 8px;
-        border-radius: 4px;
-        margin: 4px;
-        font-size: 0.875rem;
+    /* Status indicators */
+    .status-active {
+        color: #22c55e !important;  /* green-500 */
+        font-weight: 500 !important;
+    }
+    
+    .status-completed {
+        color: #94a3b8 !important;  /* slate-400 */
+    }
+    
+    /* Headers */
+    .dashboard-header {
+        color: #f1f5f9 !important;  /* slate-100 */
+        font-size: 1.5rem !important;
+        font-weight: 600 !important;
+        margin-bottom: 1rem !important;
     }
     
-    .feedback-active {
-        background: #ebf8ff;
-        color: #2b6cb0;
+    /* Tables */
+    table {
+        background-color: #1e293b !important;  /* slate-800 */
+        color: #f1f5f9 !important;  /* slate-100 */
     }
     
-    .feedback-inactive {
-        background: #f7fafc;
-        color: #718096;
+    th, td {
+        border-color: #334155 !important;  /* slate-700 */
     }
     """