Initial commit

skills/mxcp-expert/assets/project-templates/covid_owid/.gitignore

@@ -0,0 +1,4 @@
+
+target/
+dbt_packages/
+logs/

skills/mxcp-expert/assets/project-templates/covid_owid/README.md

@@ -0,0 +1,149 @@
+# COVID-19 OWID Example
+
+This example demonstrates how to use MXCP to create a COVID-19 data analysis API. It shows how to:
+- Fetch and cache COVID-19 data from Our World in Data (OWID)
+- Transform data using dbt and DuckDB
+- Create a natural language interface for data exploration using generic SQL tools
+
+## Features
+
+- **Comprehensive Data**: Global COVID-19 statistics from OWID
+- **Data Transformation**: dbt models for efficient querying
+- **Natural Language**: LLM-friendly query interface (prompt only)
+
+## Getting Started
+
+### Prerequisites
+
+Make sure you have the required tools installed:
+```bash
+# Install MXCP and dependencies
+pip install mxcp dbt-core dbt-duckdb
+
+# Option: Install in development mode
+cd /path/to/mxcp
+python -m venv .venv && source .venv/bin/activate
+pip install -e .
+```
+
+### Running the Example
+
+1. Navigate to the COVID example:
+   ```bash
+   cd examples/covid_owid
+   ```
+
+2. Initialize the data:
+   ```bash
+   dbt deps
+   dbt run
+   ```
+
+3. Start the MCP server:
+   ```bash
+   mxcp serve
+   ```
+
+## 🔌 Claude Desktop Integration
+
+To use this example with Claude Desktop:
+
+### 1. Locate Claude's Configuration
+
+Find your Claude Desktop configuration file:
+- **macOS**: `~/Library/Application Support/Claude/claude_desktop_config.json`
+- **Windows**: `%APPDATA%\Claude\claude_desktop_config.json`
+
+### 2. Configure the MCP Server
+
+Add this configuration to your `claude_desktop_config.json`:
+
+#### If you installed MXCP globally:
+```json
+{
+  "mcpServers": {
+    "covid": {
+      "command": "mxcp",
+      "args": ["serve", "--transport", "stdio"],
+      "cwd": "/absolute/path/to/mxcp/examples/covid_owid"
+    }
+  }
+}
+```
+
+#### If you're using a virtual environment:
+```json
+{
+  "mcpServers": {
+    "covid": {
+      "command": "/bin/bash",
+      "args": [
+        "-c",
+        "cd /absolute/path/to/mxcp/examples/covid_owid && source ../../.venv/bin/activate && mxcp serve --transport stdio"
+      ]
+    }
+  }
+}
+```
+
+**Important**: Replace `/absolute/path/to/mxcp` with the actual path to your MXCP installation.
+
+### 3. Restart Claude Desktop
+
+After saving the configuration, restart Claude Desktop to load the new MCP server.
+
+### 4. Test the Integration
+
+In Claude Desktop, try asking:
+- "Show me COVID-19 cases in the United States for 2022"
+- "Compare vaccination rates between France and Germany"
+- "What were the peak hospitalization rates in the UK?"
+
+## 🛠️ Other MCP Clients
+
+This example works with any MCP-compatible tool:
+- **mcp-cli**: Interactive command-line interface
+- **Custom integrations**: Build your own using the MCP specification
+
+## Example Usage
+
+The LLM can help you analyze:
+- Case numbers and death rates
+- Vaccination progress
+- Hospital occupancy
+- Regional comparisons
+- Policy effectiveness
+
+All queries are handled through the generic SQL query interface. You can:
+- Use `list_tables` to see available tables
+- Use `get_table_schema` to inspect table structure
+- Use `execute_sql_query` to run custom SQL queries
+
+## Implementation Details
+
+The example uses:
+- dbt for data transformation
+- DuckDB for efficient storage and querying
+- SQL analytics for complex calculations
+- Type-safe parameters for filtering
+
+## Project Structure
+
+```
+covid_owid/
+├── endpoints/                # MCP endpoint definitions
+│   └── prompt.yml           # LLM system prompt (generic query interface only)
+├── models/                  # dbt transformations
+│   ├── covid_data.sql      # Main COVID-19 statistics
+│   ├── hospitalizations.sql # Hospital/ICU data
+│   └── locations.sql       # Geographic data
+├── mxcp-site.yml           # MCP configuration
+└── dbt_project.yml         # dbt configuration
+```
+
+## Learn More
+
+- [OWID COVID-19 Data](https://github.com/owid/covid-19-data) - Data source
+- [dbt Documentation](https://docs.getdbt.com/) - Data transformation
+- [DuckDB Documentation](https://duckdb.org/docs/) - Database engine
+- [MXCP Documentation](../../docs/quickstart.md) - MCP framework

skills/mxcp-expert/assets/project-templates/covid_owid/dbt_project.yml

@@ -0,0 +1,19 @@
+analysis-paths:
+- analyses
+clean-targets:
+- target
+- dbt_packages
+config-version: 2
+macro-paths:
+- macros
+model-paths:
+- models
+name: covid_owid
+profile: covid_owid_default
+seed-paths:
+- seeds
+snapshot-paths:
+- snapshots
+target-path: target
+test-paths:
+- tests

skills/mxcp-expert/assets/project-templates/covid_owid/models/covid_data.sql

@@ -0,0 +1,4 @@
+{{ config(materialized='table') }}
+
+select *
+from read_csv_auto('https://github.com/owid/covid-19-data/raw/master/public/data/owid-covid-data.csv')

skills/mxcp-expert/assets/project-templates/covid_owid/models/hospitalizations.sql

@@ -0,0 +1,5 @@
+{{ config(materialized='table') }}
+
+select *
+
+from read_csv_auto('https://github.com/owid/covid-19-data/raw/master/public/data/hospitalizations/covid-hospitalizations.csv')

skills/mxcp-expert/assets/project-templates/covid_owid/models/locations.sql

@@ -0,0 +1,5 @@
+{{ config(materialized='table') }}
+
+select *
+
+from read_csv_auto('https://github.com/owid/covid-19-data/raw/master/public/data/hospitalizations/locations.csv')

skills/mxcp-expert/assets/project-templates/covid_owid/models/sources.yml

@@ -0,0 +1,9 @@
+version: 2
+
+sources:
+  - name: github_covid_data
+    description: "COVID-19 data loaded directly from Our World in Data GitHub repository"
+    meta:
+      urls:
+        owid_covid_data: "https://github.com/owid/covid-19-data/raw/master/public/data/owid-covid-data.csv"
+        covid_hospitalizations: "https://github.com/owid/covid-19-data/raw/master/public/data/hospitalizations/covid-hospitalizations.csv"

skills/mxcp-expert/assets/project-templates/covid_owid/mxcp-site.yml

@@ -0,0 +1,5 @@
+mxcp: 1
+profile: default
+project: covid_owid
+sql_tools:
+  enabled: true

skills/mxcp-expert/assets/project-templates/covid_owid/prompts/prompt.yml

@@ -0,0 +1,75 @@
+mxcp: 1
+
+prompt:
+  name: "covid_data_analyst"
+  description: "An AI assistant that analyzes and explains COVID-19 data from Our World in Data."
+  tags: ["covid", "analysis", "health", "epidemiology"]
+  messages:
+    - role: system
+      type: text
+      prompt: |
+        You are an expert COVID-19 data analyst with access to the Our World in Data (OWID) COVID-19 dataset. You can help users understand and analyze:
+
+        1. Case numbers, deaths, and testing data
+        2. Vaccination rates and their impact
+        3. Hospital and ICU occupancy rates
+        4. Regional and country-specific trends
+        5. Comparative analysis between countries
+        6. Policy responses and their effectiveness
+
+        Data Exploration Tools:
+        You have access to a generic query interface for exploring the COVID-19 data:
+        - list_tables: View all available tables in the database
+        - get_table_schema: Examine the structure and columns of any table
+        - execute_sql_query: Run custom SQL queries for data analysis
+
+        These tools allow you to:
+        1. Explore available data tables and their structure
+        2. Create custom queries for specific analysis needs
+        3. Perform complex aggregations and calculations
+        4. Combine data from different tables
+        5. Filter and sort data in any way needed
+        6. Answer detailed or unusual questions from users
+
+        Available data includes:
+        - Daily and cumulative case counts
+        - Death rates and mortality statistics
+        - Testing rates and positivity rates
+        - Vaccination data (first, second doses, boosters)
+        - Hospital and ICU admissions
+        - Demographics and population metrics
+        - Government response indicators
+
+        When responding:
+        - Use list_tables and get_table_schema to understand available data
+        - Create focused SQL queries that answer the specific question
+        - Provide context for the numbers you present
+        - Explain trends and potential factors affecting the data
+        - Note any data limitations or gaps
+        - Use clear, non-technical language when possible
+        - Cite specific dates and sources
+        - Acknowledge uncertainty where it exists
+        - For SQL queries, explain your logic
+
+        Example Usage:
+        1. Explore available tables:
+           list_tables()
+        
+        2. Understand table structure:
+           get_table_schema("covid_data")
+        
+        3. Custom analysis:
+           execute_sql_query("
+             SELECT 
+               location, 
+               date, 
+               new_cases,
+               new_deaths,
+               total_vaccinations
+             FROM covid_data 
+             WHERE date >= '2021-01-01'
+             AND location IN ('United States', 'United Kingdom')
+             ORDER BY date DESC
+           ")
+
+        The data is sourced from Our World in Data's COVID-19 dataset, which is regularly updated and maintained by researchers at the University of Oxford.