quick-start-integration.md•2.81 kB
# Quick Start: Creating Custom Integrations
This guide provides a quick introduction to creating custom integrations with Createve.AI Nexus. For complete details, see the [Custom API Guide](custom-api-guide.md).
## Overview
Createve.AI Nexus makes it easy to expose your custom functionality as both REST API endpoints and MCP tools. Each integration is automatically available to:
- Microsoft Copilot Studio agents
- Anthropic's Claude (via desktop app)
- Any OpenAPI-compatible client
- Any MCP-enabled system
## Basic Integration Example
Here's a simple example that exposes enterprise data analysis capabilities:
```python
class DataAnalyzer:
"""Enterprise data analysis endpoint."""
CATEGORY = "analytics"
@classmethod
def INPUT_TYPES(cls):
return {
"required": {
"query": ("STRING", {"multiline": True}),
"dataset": ("STRING", {"default": "production"})
},
"optional": {
"time_range": ("STRING", {"default": "1d"})
}
}
RETURN_TYPES = ("DICT",)
RETURN_NAMES = ("analysis_results",)
FUNCTION = "analyze_data"
def analyze_data(self, query, dataset, time_range="1d"):
# Implementation here
return (result,)
```
## Key Concepts
1. **Class Structure**
- Clear docstring describing functionality
- Category for grouping in documentation
- Input types definition with required/optional parameters
- Return type specifications
- Main processing function
2. **Automatic Exposure**
- REST API endpoint at `/api/analytics/dataAnalyzer`
- MCP tool named `analytics_dataAnalyzer`
- OpenAPI documentation automatically generated
3. **Type System**
- Rich type system for inputs/outputs
- Automatic validation and conversion
- Support for files, images, and complex data
## Quick Start Steps
1. Create a Python file in the `custom_apis` directory
2. Define your class following the example structure
3. Implement your processing logic
4. Restart the server to load your new integration
## Using Your Integration
### REST API
```bash
curl -X POST "http://localhost:43080/api/analytics/dataAnalyzer" \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{"query": "sales_by_region", "dataset": "production"}'
```
### MCP (via Copilot Studio or Claude)
Your integration automatically appears as an available tool in MCP-enabled clients.
### OpenAPI
Browse to `http://localhost:43080/docs` to see your endpoint in the Swagger UI.
## Next Steps
- Read the [Custom API Guide](custom-api-guide.md) for detailed documentation
- Explore example integrations in the `custom_apis` directory
- Check the [MCP Integration Guide](mcp-integration.md) for advanced MCP features