Digital Archaeology Mystery

📦 Step 1: Install Ollama

Download and install Ollama from https://ollama.ai

            macOS/Linux:

            curl -fsSL https://ollama.ai/install.sh | sh

            Windows:

            Download the installer from the website

🔧 Step 2: Configure CORS for Web Access

To allow this website (https://jtm.io/codepedagogy/) to connect to your local Ollama server, you need to set environment variables:

Windows Setup:

              Option 1 - Command Prompt:

              set OLLAMA_ORIGINS=https://jtm.io/codepedagogy/,http://localhost:*

              ollama serve

              Option 2 - PowerShell:

              $env:OLLAMA_ORIGINS="https://jtm.io/codepedagogy/,http://localhost:*"

              ollama serve

              Option 3 - Permanent (System Environment Variables):

              1. Press Win + R, type sysdm.cpl

              2. Click "Environment Variables"

              3. Add new system variable:

                Name: OLLAMA_ORIGINS

                Value: https://jtm.io/codepedagogy/,http://localhost:*

              4. Restart command prompt and run ollama serve

🚀 Step 3: Install a Model

Download a model suitable for coding assistance:

            Recommended models:

            ollama pull codellama:7b (good for coding, 4GB)

            ollama pull llama3.1:8b (general purpose, 5GB)

            ollama pull qwen2.5-coder:7b (excellent for coding, 4GB)

✅ Step 4: Test Connection

Once Ollama is running with CORS configured:

Enable the "🤖 AI Assistant" toggle above
Select your model from the dropdown
Look for "Connected" status

⚠️ Important Security Notes

This configuration allows the specific website to access your local Ollama instance
Only use this setup on trusted networks
Stop Ollama when not needed: Ctrl+C in the terminal
Your models and conversations stay completely private on your machine

🔍 Troubleshooting

Connection failed: Ensure Ollama is running and CORS is configured
Model not found: Make sure you've pulled the model with ollama pull model-name
Slow responses: Try a smaller model or check your system resources
Windows issues: Try running Command Prompt as Administrator

🏛️ Digital Archaeology Mystery

The Mystery

Your Code