API_REFERENCE.md•5.62 kB
# API Reference
This document provides detailed specifications for all tools exposed by the Analytical MCP Server.
## Tool Specifications
All tools are accessed with the `analytical:` prefix when used through Claude Desktop (e.g., `analytical:analyze_dataset`).
### analytical:analyze_dataset
Statistical analysis of datasets.
**Parameters:**
- `data` (array): Array of numeric values or objects
- `analysisType` (string): "summary" or "stats" (default: "summary")
**Returns:** Statistical analysis including mean, median, standard deviation, quartiles.
**Example:**
```json
{
"data": [1, 2, 3, 4, 5, 6, 7, 8, 9, 10],
"analysisType": "stats"
}
```
### analytical:decision_analysis
Multi-criteria decision analysis with weighted scoring.
**Parameters:**
- `options` (array): Decision options to analyze
- `criteria` (array): Evaluation criteria
- `weights` (array, optional): Criterion weights (must match criteria length)
**Returns:** Ranked options with scores and analysis.
**Example:**
```json
{
"options": ["Option A", "Option B", "Option C"],
"criteria": ["Cost", "Quality", "Speed"],
"weights": [0.4, 0.4, 0.2]
}
```
### analytical:advanced_regression_analysis
Regression analysis on datasets.
**Parameters:**
- `data` (array): Data objects for regression
- `regressionType` (string): "linear", "polynomial", "logistic", or "multivariate"
- `independentVariables` (array): Independent variable names
- `dependentVariable` (string): Dependent variable name
**Returns:** Regression equation, coefficients, R-squared, model statistics.
**Example:**
```json
{
"data": [
{"sales": 1200, "advertising": 100},
{"sales": 1500, "advertising": 150},
{"sales": 1800, "advertising": 200}
],
"regressionType": "linear",
"independentVariables": ["advertising"],
"dependentVariable": "sales"
}
```
### analytical:hypothesis_testing
Statistical hypothesis testing.
**Parameters:**
- `testType` (string): "t_test_independent", "t_test_paired", "correlation", "chi_square", "anova"
- `data` (array): Test data (format depends on test type)
- `variables` (array, optional): Variable names for complex data
- `alpha` (number, optional): Significance level (default: 0.05)
- `alternativeHypothesis` (string, optional): Alternative hypothesis
**Returns:** Test statistics, p-values, conclusions.
**Example:**
```json
{
"testType": "t_test_independent",
"data": [[23, 45, 67], [34, 56, 78]],
"alpha": 0.05
}
```
### analytical:data_visualization_generator
Data visualization specifications.
**Parameters:**
- `data` (array): Data objects to visualize
- `visualizationType` (string): "scatter", "line", "bar", "histogram", "box", "heatmap", "pie"
- `variables` (array): Variable names to include
**Returns:** Visualization specification with chart configuration.
### analytical:logical_argument_analyzer
Logical argument analysis for structure, fallacies, validity, and strength.
**Parameters:**
- `argument` (string): Argument text to analyze
- `analysisDepth` (string, optional): "basic" or "comprehensive" (default: "basic")
**Returns:** Argument structure analysis, validity assessment, improvement recommendations.
### analytical:logical_fallacy_detector
Logical fallacy detection in text.
**Parameters:**
- `text` (string): Text to analyze
- `confidenceThreshold` (number, optional): Minimum confidence level (0-1, default: 0.7)
**Returns:** Detected fallacies with explanations and confidence scores.
### analytical:perspective_shifter
Alternative perspectives on problems or situations.
**Parameters:**
- `problem` (string): Problem or situation to analyze
- `currentPerspective` (string, optional): Current perspective or viewpoint
- `shiftType` (string, optional): "stakeholder", "discipline", "contrarian"
- `numberOfPerspectives` (number, optional): Number of perspectives (default: 3)
**Returns:** Alternative perspectives with analysis and insights.
### analytical:verify_research
Research claim verification from multiple sources.
**Parameters:**
- `query` (string): Primary research query
- `verificationQueries` (array, optional): Alternative queries for verification
- `minConsistencyThreshold` (number, optional): Minimum consistency score (0-1, default: 0.7)
- `sources` (number, optional): Number of sources to verify (1-10, default: 3)
**Returns:** Verification results with consistency scores and source analysis.
## Error Handling
Tools return structured error messages for:
- Invalid parameters
- Missing required data
- External API failures (research tools)
- Processing errors
Error responses include:
- Error type and code
- Error message
- Resolution suggestions
## Rate Limiting
Research tools using external APIs:
- Built-in retry logic with exponential backoff
- Configurable timeout settings
- Graceful degradation when limits are reached
## Logging System
### Logger Implementation
- Centralized Logger class in src/utils/logger.ts
- MCP protocol compliance (stderr for logs, stdout for protocol communication)
- Singleton pattern for consistent instance usage
- Log levels: info, warn, error
### Usage Pattern
```typescript
import { Logger } from './utils/logger.js';
const logger = Logger.getInstance();
logger.info('Information message');
logger.warn('Warning message');
logger.error('Error message');
```
### Utility Script Integration
- tools/cache-manager.js: Cache management with Logger integration
- tools/check-api-keys.js: API key validation with Logger integration
- All utility scripts use Logger instead of console statements
- Examples directory preserves console output for demonstration clarity