lineai-mcp-server copilot-instructions.md

Lineai MCP Server - AI Assistant Instructions

About This Codebase

This repository contains a Model Context Protocol (MCP) server that integrates with Lineai's knowledge graph APIs. It enables AI programming assistants (like GitHub Copilot and Cursor) to access dependency data from Lineai to analyze code and database impacts.

Key Technologies

• Python 3.13+ with extensive use of async/await
• Model Context Protocol SDK (mcp[cli])
• Neo4j for graph database operations
• HTTPX for API requests
• Environment variables via dotenv for configuration

Project Structure

• src/lineai_mcp_server/: Core package
• __init__.py: Package initialization and entry point
• server.py: MCP server implementation
• handlers.py: Tool handlers implementation
• utils.py: API interaction utilities

Core Coding Patterns

MCP Server Pattern

server = Server("lineai-mcp-server")

@server.list_tools()
async def handle_list_tools() -> list[types.Tool]:
    # Define and return tools
    
@server.call_tool()
async def handle_call_tool(name: str, arguments: dict | None) -> list[types.TextContent]:
    # Handle tool execution

Error Handling

try:
    # Operations that might fail
except Exception as e:
    sys.stderr.write(f"Error: {str(e)}\n")
    return [types.TextContent(type="text", text=f"# Error\n\n{str(e)}")]

Style Guidelines

1. Copyright Headers: Include MPL 2.0 headers in all Python files
2. Docstrings: Google-style docstrings for modules/classes/functions
3. Type Hints: Always use Python type hints
4. Asynchronous: Keep I/O operations asynchronous
5. Format Outputs: Return markdown-formatted text in tool responses

Tool Implementation Pattern

When implementing new MCP tools:

1. Add to handle_list_tools() with descriptive name (prefix: lineai-)
2. Add handler in handle_call_tool()
3. Implement handler function with error handling
4. Return results as markdown-formatted text