# Toronto MCP Usage Examples
This document shows examples of how to use the advanced Toronto Open Data MCP tools from AI assistants like Claude, Cursor, or other MCP clients.
## Scenario 1: Research Assistant - Finding Traffic Data
**User Query**: "I need to analyze traffic accident patterns in Toronto. What data is available?"
**MCP Tool Usage**:
```json
{
"tool": "find_relevant_datasets",
"parameters": {
"query": "traffic accidents collisions",
"maxResults": 5,
"includeRelevanceScore": true
}
}
```
**Expected Response**: Ranked list of datasets like "Motor Vehicle Collisions", "Traffic Signals", with relevance scores and update frequencies.
## Scenario 2: Data Quality Assessment
**User Query**: "How often does Toronto update their parking and transportation data?"
**MCP Tool Usage**:
```json
{
"tool": "analyze_dataset_updates",
"parameters": {
"query": "parking transportation",
"groupByFrequency": true
}
}
```
**Expected Response**: Categorized breakdown showing which datasets update daily, weekly, monthly, etc.
## Scenario 3: Data Structure Analysis
**User Query**: "What fields are available in Toronto's building permits dataset?"
**MCP Tool Usage**:
```json
{
"tool": "analyze_dataset_structure",
"parameters": {
"packageId": "building-permits",
"includeDataPreview": true,
"previewLimit": 5
}
}
```
**Expected Response**: Field definitions, data types, record counts, and sample data.
## Scenario 4: Comprehensive Research
**User Query**: "Give me insights about Toronto's housing and development data"
**MCP Tool Usage**:
```json
{
"tool": "get_dataset_insights",
"parameters": {
"query": "housing development permits construction",
"maxDatasets": 3,
"includeUpdateFrequency": true,
"includeDataStructure": true
}
}
```
**Expected Response**: Complete analysis including:
- Most relevant datasets
- Update frequency analysis
- Data structure overview
- Suggested related searches
## Scenario 5: Exploring Data Categories
**User Query**: "What categories of data does Toronto provide?"
**MCP Tool Usage**:
```json
{
"tool": "get_data_categories",
"parameters": {}
}
```
**Expected Response**: List of all organizations (city departments) and topic groups with dataset counts.
## Advanced Query Examples
### Complex Search with Filters
Find datasets about environment and sustainability:
```json
{
"tool": "find_relevant_datasets",
"parameters": {
"query": "environment sustainability green energy",
"maxResults": 10,
"includeRelevanceScore": true
}
}
```
### Update Pattern Analysis for Specific Datasets
Analyze specific datasets by ID:
```json
{
"tool": "analyze_dataset_updates",
"parameters": {
"packageIds": ["air-quality", "water-quality", "green-space"],
"groupByFrequency": true
}
}
```
### Deep Data Analysis
Get comprehensive structure analysis with data preview:
```json
{
"tool": "analyze_dataset_structure",
"parameters": {
"packageId": "financial-performance",
"includeDataPreview": true,
"previewLimit": 10
}
}
```
## AI Assistant Integration Tips
1. **Chain Multiple Tools**: Start with `find_relevant_datasets`, then use `analyze_dataset_structure` for detailed analysis.
2. **Update Frequency Awareness**: Use `analyze_dataset_updates` to inform users about data freshness.
3. **Smart Recommendations**: Use `get_dataset_insights` for comprehensive responses to broad queries.
4. **Category Exploration**: Use `get_data_categories` to help users discover what's available.
5. **Progressive Disclosure**: Start with basic search, then drill down based on user interest.
## Error Handling
The tools are designed to be robust:
- Invalid dataset IDs return helpful error messages
- Missing datastore resources are handled gracefully
- Network timeouts have appropriate fallbacks
- Large datasets are paginated automatically
## Performance Notes
- `find_relevant_datasets` uses smart caching and scoring
- `analyze_dataset_structure` can be expensive for datasets with many resources
- `get_dataset_insights` combines multiple API calls efficiently
- All tools support parameter validation via Zod schemas
