5.Add an MCP

class DatabaseMCP:
    """
    MCP server for SQLite database operations.
    Provides tools for creating tables, inserting data, and querying databases.
    """
    command = "uvx"
    args = ["mcp-server-sqlite", "--db-path", "/path/to/database.db"]

class MCP:
    # Missing documentation and unclear purpose
    command = "uvx"
    args = ["some-server"]

class SecureMCP:
    """
    MCP server with security considerations.
    """
    command = "uvx"
    args = ["mcp-server-sqlite", "--db-path", "/secure/path/database.db"]
    env = {"SECURE_MODE": "true"}  # Environment variables for security

class WellDocumentedMCP:
    """
    Clear description of what the MCP does.
    
    This MCP provides filesystem operations including:
    - Reading files
    - Writing files
    - Directory listing
    - File permissions management
    
    Security Note: Only operates within the specified directory.
    """
    command = "uvx"
    args = ["mcp-server-filesystem", "/allowed/directory"]

# Upsonic Docs: Add an MCP
# https://docs.upsonic.ai/guides/5-add-an-mcp

# Imports
import os
import tempfile
from upsonic import Agent, Task
from upsonic.tools import WebSearch, WebRead
from pydantic import BaseModel

# Set your OpenAI API key
os.environ["OPENAI_API_KEY"] = "api_key"

# MCP 1: SQLite Database Operations
class DatabaseMCP:
    """
    MCP server for SQLite database operations.
    Provides comprehensive database management capabilities.
    """
    command = "uvx"
    args = ["mcp-server-sqlite", "--db-path", "/tmp/library.db"]

# Response Format for Structured Output
class DatabaseReport(BaseModel):
    tables_created: int
    records_inserted: int
    sample_data: str
    summary: str

# Agent Creation with MCP Tools
database_agent = Agent(
    name="Database Management Expert",
    role="Database operations and data management specialist",
    goal="Provide comprehensive database management with structured reporting"
)

# Example 1: Basic Database Operations
basic_task = Task(
    description="Create a comprehensive library management database. Create a 'books' table with columns: id (integer primary key), title (text), author (text), isbn (text), published_year (integer), and available (boolean). Insert 5 sample books with different authors and years. Then create a 'borrowers' table with columns: id (integer primary key), name (text), email (text), and phone (text). Insert 3 sample borrowers. Finally, show all books and all borrowers in separate queries.",
    tools=[DatabaseMCP]
)

# Execute the basic database operations
result = database_agent.do(basic_task)
print("Basic Database Operations Result:")
print(result)

# Example 2: Structured Output with MCP
structured_task = Task(
    description="Create a 'products' table with id, name, price, and category columns. Insert 5 sample products with different categories. Then provide a structured report including the number of tables created, records inserted, sample data, and a summary of the operations.",
    tools=[DatabaseMCP],
    response_format=DatabaseReport
)

# Execute with structured output
structured_result = database_agent.do(structured_task)
print("\nStructured Database Report:")
print(f"Tables Created: {structured_result.tables_created}")
print(f"Records Inserted: {structured_result.records_inserted}")
print(f"Summary: {structured_result.summary}")

# Example 3: Combined MCP and Regular Tools
research_agent = Agent(
    name="Research and Database Agent",
    role="Combined research and data management specialist",
    goal="Research topics and store findings in a structured database"
)

combined_task = Task(
    description="Research the top 3 programming languages in 2024. Create a 'languages' table with columns: id, name, popularity_rank, description, and use_cases. Insert the research findings into the database. Then query the database to show all languages sorted by popularity rank.",
    tools=[WebSearch, WebRead, DatabaseMCP]
)

# Execute the combined workflow
combined_result = research_agent.do(combined_task)
print("\nCombined Research and Database Result:")
print(combined_result)

# Example 4: Multiple MCP Workflow
workflow_agent = Agent(
    name="Multi-MCP Workflow Agent",
    role="Multi-tool specialist",
    goal="Use multiple MCP tools to accomplish complex tasks"
)

workflow_task = Task(
    description="Create a 'companies' table with id, name, industry, and revenue columns. Research 3 tech companies and insert their information into the database. Then create a 'employees' table and insert sample employee data for these companies. Finally, provide a comprehensive report of all data.",
    tools=[WebSearch, WebRead, DatabaseMCP]
)

# Execute the workflow
workflow_result = workflow_agent.do(workflow_task)
print("\nMulti-MCP Workflow Result:")
print(workflow_result)