Add Modal deployment setup guide

MODAL_SETUP.md

@@ -0,0 +1,123 @@
+# 🏛️ HERMES Modal Deployment Setup
+
+## ✅ Prerequisites Completed
+
+- ✅ Python 3.13 virtual environment created (`.venv/`)
+- ✅ Modal CLI installed (version 1.2.2)
+- ✅ MCP server code ready ([mcp_server.py](mcp_server.py))
+
+## 📋 Next Steps - Authentication Required
+
+### 1. Authenticate with Modal
+
+Run this command to set up your Modal account:
+
+```bash
+source .venv/bin/activate
+modal token new
+```
+
+This will:
+- Open a browser window
+- Prompt you to sign up/sign in to Modal (it's free!)
+- Create API credentials on your machine
+- Link your local environment to your Modal workspace
+
+### 2. Verify Authentication
+
+After authenticating, check that it worked:
+
+```bash
+source .venv/bin/activate
+modal profile list
+```
+
+You should see your workspace name in the table.
+
+### 3. Deploy HERMES MCP Server
+
+Once authenticated, deploy the server:
+
+```bash
+source .venv/bin/activate
+modal deploy mcp_server.py
+```
+
+This will:
+- Package your code and dependencies
+- Deploy to Modal's serverless infrastructure
+- Give you a live URL for your MCP functions
+
+### 4. Test the Deployment
+
+Test your deployed functions locally:
+
+```bash
+source .venv/bin/activate
+modal run mcp_server.py
+```
+
+This runs the local test suite that calls your deployed functions.
+
+### 5. (Optional) Set Up Secrets for AI Features
+
+If you want to use OpenAI/Anthropic/ElevenLabs features in the MCP server:
+
+```bash
+source .venv/bin/activate
+modal secret create hermes-secrets \
+  OPENAI_API_KEY=your-openai-key \
+  ANTHROPIC_API_KEY=your-anthropic-key \
+  ELEVENLABS_API_KEY=your-elevenlabs-key
+```
+
+Then update [mcp_server.py](mcp_server.py) to use the secrets.
+
+## 🎯 What Modal Gives You
+
+- **Serverless Functions**: Your astrology calculations run on-demand
+- **Automatic Scaling**: Handles 1 user or 1000 users seamlessly
+- **Free Tier**: 30 free credits/month (plenty for development)
+- **MCP Integration**: Functions are exposed as MCP tools
+- **No Infrastructure**: No servers to manage
+
+## 📊 Available MCP Functions
+
+Once deployed, these functions will be available:
+
+1. `calculate_full_dignities` - Complete essential dignity analysis
+2. `get_bound_ruler` - Egyptian bounds (terms) calculation
+3. `get_decan_ruler` - Decan (face) calculation
+4. `calculate_zodiacal_releasing` - Time-lord periods from Vettius Valens
+5. `calculate_firdaria` - Persian time-lord system
+6. `find_fixed_stars` - Fixed star conjunctions
+
+## 💰 Cost Estimate
+
+Based on the current configuration:
+- **CPU**: 1 vCPU @ 0.1 credits/hour
+- **Memory**: 1GB RAM (included)
+- **Typical Request**: ~0.5 seconds = ~0.00001 credits
+- **Free Tier**: 30 credits/month ≈ 300,000 requests
+
+## 🔗 Resources
+
+- **Modal Docs**: https://modal.com/docs
+- **Modal Dashboard**: https://modal.com/apps
+- **Billing**: https://modal.com/settings/billing
+- **Support**: https://modal.com/slack
+
+## 🐛 Troubleshooting
+
+**Problem**: `modal token new` doesn't work
+- **Solution**: Make sure you're in the virtual environment (`source .venv/bin/activate`)
+
+**Problem**: Deployment fails
+- **Solution**: Check [mcp_server.py](mcp_server.py) syntax with `python mcp_server.py`
+
+**Problem**: Functions timeout
+- **Solution**: Increase timeout in `modal_config.py` (currently 300 seconds)
+
+---
+
+**Ready to deploy?** Start with step 1 above! 🚀