"""
Data analysis prompts for text processing and classification.
These prompts demonstrate FastMCP 2.x decorator-based prompt patterns with:
- Simplified Field() pattern for parameter validation
- Type hints with parameterized types (dict[str, str])
- Docstrings for documentation
- Optional parameters with defaults
- FastMCP 2.9.0+ compatibility
"""
from pydantic import Field
# Import the shared mcp instance from core
from ...core.app import mcp
@mcp.prompt
def summarize(
document: str = Field(
description="The document text to summarize",
min_length=1,
),
) -> str:
"""
Summarize a document with clear section markers.
This prompt takes a document and requests a concise summary with key points.
The response should be structured JSON with a summary field and key_points array.
Args:
document: The text content to be summarized
Returns:
A formatted prompt string requesting a structured summary
"""
return f"""Summarize the following text:
<document>{document}</document>
Use clear, concise language.
Return your response as JSON with this structure:
{{
"summary": "A concise summary of the main content",
"key_points": ["Key point 1", "Key point 2", "Key point 3"]
}}"""
@mcp.prompt
def classify(
text: str = Field(
description="The text to classify into categories",
min_length=1,
),
) -> str:
"""
Classify text into categories with confidence scores.
This prompt analyzes text and assigns it to appropriate categories
along with confidence scores for the classification.
Args:
text: The text content to classify
Returns:
A formatted prompt string requesting structured classification results
"""
return f"""Classify the following text:
<text>{text}</text>
Return JSON matching this schema:
{{
"category": "The primary category for this text",
"confidence": 0.95
}}"""
@mcp.prompt
def analyze_sentiment(
text: str = Field(
description="The text to analyze for sentiment",
min_length=1,
),
) -> str:
"""
Analyze the sentiment of given text (positive, negative, neutral).
Args:
text: The text content to analyze
Returns:
A formatted prompt string requesting sentiment analysis
"""
return f"""Analyze the sentiment of the following text:
<text>{text}</text>
Provide:
1. Overall sentiment (positive, negative, or neutral)
2. Sentiment score (-1.0 to 1.0)
3. Key phrases that influenced the sentiment
4. Brief explanation of the analysis
Return as JSON:
{{
"sentiment": "positive|negative|neutral",
"score": 0.8,
"key_phrases": ["phrase1", "phrase2"],
"explanation": "Brief explanation"
}}"""
@mcp.prompt
def extract_entities(
text: str = Field(
description="The text to extract named entities from",
min_length=1,
),
entity_types: list[str] | None = Field(
default=None,
description=(
"Specific entity types to extract "
"(e.g., ['PERSON', 'ORGANIZATION', 'LOCATION'])"
),
),
) -> str:
"""
Extract named entities from text with optional filtering by entity type.
This example shows how to handle optional parameters in prompts.
Args:
text: The text content to analyze
entity_types: Optional list of specific entity types to extract
Returns:
A formatted prompt string requesting entity extraction
"""
entity_filter = ""
if entity_types:
types_str = ", ".join(entity_types)
entity_filter = f"\nFocus on these entity types: {types_str}"
return f"""Extract named entities from the following text:
<text>{text}</text>{entity_filter}
Return JSON with entities grouped by type:
{{
"PERSON": ["John Doe", "Jane Smith"],
"ORGANIZATION": ["ACME Corp", "Tech Inc"],
"LOCATION": ["New York", "San Francisco"],
"DATE": ["2024-01-01"],
"OTHER": ["Any other relevant entities"]
}}"""
@mcp.prompt
def analyze_data(
data: dict[str, str] = Field(
description=(
"Data to analyze as a dictionary with string keys and values. "
"Pass as JSON string."
)
),
analysis_type: str = Field(
default="summary",
description="Type of analysis: 'summary', 'detailed', or 'statistical'",
),
) -> str:
"""
Analyze structured data and provide insights.
This example demonstrates FastMCP 2.9.0+ dict parameter handling.
Clients pass dict as JSON string, FastMCP converts automatically.
Args:
data: Dictionary of data to analyze
analysis_type: Type of analysis to perform
Returns:
Formatted prompt for data analysis
"""
return f"""Analyze the following data:
{data}
Perform a {analysis_type} analysis including:
1. Key patterns and trends
2. Notable outliers or anomalies
3. Statistical summaries (if applicable)
4. Actionable insights
Return results as structured JSON."""
