Quantum ZX-Calculus MCP Server

# Quantum ZX-Calculus MCP - Standard Structure Pattern This project follows the **Standard MCP Server Setup Pattern** established by Lushy. ## Phase 1: Automated Structure Generation Run `create_structure.sh` to generate: - Directory hierarchy (src/, tests/, docs/, data/) - __init__.py files - .gitignore - __main__.py (local execution) - handler.py (FastMCP Cloud) - README.md - pyproject.toml - This documentation ## Phase 2: Manual Large File Placement Copy these files to the project root: - `quantum_zx_ologs.py` - Taxonomy definitions (gates, spiders, rules) - `quantum_zx_calculus.py` - Core ZX-calculus implementation - `quantum_zx_server.py` - FastMCP server definition - `test_quantum_zx.py` - Comprehensive tests ## Phase 3: Verification Run `verify_structure.sh` to validate: 1. Directory structure complete 2. All required files present 3. Import paths correct 4. Dependencies available ## Installation & Testing ```bash # Install in development mode pip install -e ".[dev]" # Run tests ./tests/run_tests.sh # Run locally python -m quantum_zx ``` ## FastMCP Cloud Deployment Entry point: `quantum_zx/handler.py:handler` The handler function returns the MCP server object. FastMCP Cloud handles the event loop and server.run() call. ## Cost Optimization Strategy - **Layer 1:** Deterministic gate taxonomy (0 tokens) - **Layer 2:** Rewrite rule composition (0 tokens) - **Layer 3:** Circuit analysis (deterministic mapping, 0 tokens) - **Layer 4:** Claude synthesis (single call, ~200-400 tokens) **Result:** 60-85% cost savings vs pure LLM approach