CoreMCPKit

# CoreMCPKit **CoreMCPKit** is a modular, plugin-based Swift framework built to support the **Model Context Protocol (MCP)**. It enables developers to create context-aware AI agents that integrate seamlessly with various Large Language Models (LLMs) and dynamic data sources. --- ## 🚀 Features - 🧠 **LLM-Agnostic**: Compatible with OpenAI, Claude, Mistral, and more. - 🧩 **Plugin Architecture**: Contexts are managed by independent providers. - 💬 **Emotion-Aware Responses**: Detects user emotions and adapts tone dynamically. - 🔄 **Multi-Context Support**: Combines multiple contexts for richer LLM interactions. - 🔧 **Easy Configuration**: Integrate your own API keys with minimal setup. --- ## 📦 Installation ### Swift Package Manager Add **CoreMCPKit** to your project using Swift Package Manager by including it in your `Package.swift` file: ```swift dependencies: [ .package(url: "https://github.com/isamercan/CoreMCPKit.git", .upToNextMajor(from: "1.0.0")) ] ``` Then, import the library in your code: ```swift import CoreMCPKit ``` --- ## ⚙️ Configuration Configure **CoreMCPKit** by setting up the `MCPAgentManager` and registering context providers: ```swift import CoreMCPKit // Initialize configuration with your OpenAI API key let config = MCPConfiguration(openAIApiKey: "sk-...") // Create the MCP agent manager let manager = MCPAgentManager(config: config) // Register context providers let openAI = OpenAIProvider(apiKey: config.openAIApiKey) let parser = PromptToFlexibleQueryParser(openAIService: openAI) manager.registerProvider(FlexibleContextProvider(parser: parser)) manager.registerProvider(EmotionContextProvider(openAIService: openAI)) ``` --- ## 📝 Example Usage Process user input and generate context-aware responses with the following code: ```swift Task { do { let userInput = "I'm looking for affordable villas in Antalya for June, prices are too high!" let response = try await manager.respond(to: userInput) print(response) } catch { print("Error: \(error)") } } ``` --- ## Sample App A sample SwiftUI app is included in the `/Sample/CoreMCPKitSampleApp/` directory. To run: 1. Open `CoreMCPKitSampleApp.xcodeproj`. 2. Add `CoreMCPKit` as a local Swift Package. 3. Build and run the app. ## 📚 Additional Information - **Documentation**: Visit the [GitHub Wiki](https://github.com/username/CoreMCPKit/wiki) for detailed guides. - **Support**: Reach out with questions via [GitHub Issues](https://github.com/username/CoreMCPKit/issues). - **License**: CoreMCPKit is distributed under the MIT License.