Awesome Cursor MCP Server

The Fonoster MCP Server is a utility server that integrates Fonoster's open-source telephony platform with AI assistants through the Model Context Protocol (MCP). This server enables AI models to interact with Fonoster's programmable voice applications, facilitating seamless management of telephony services. Key features include:​ - List Numbers: Retrieves a list of phone numbers associated with the Fonoster account, displaying details such as references, names, and telephone URLs. - List Applications: Provides a list of voice applications configured within Fonoster, allowing for efficient management and deployment of telephony applications. Designed for local operation, the server ensures data privacy and security. It is actively maintained as part of the Fonoster open-source project.

The Cursor Talk to Figma MCP Server is a utility server that integrates the Cursor AI coding assistant with Figma through the Model Context Protocol (MCP). This integration enables AI models to access and modify Figma design files programmatically, facilitating seamless design-to-code workflows. Key features include:​ - Document and Selection Retrieval: Fetches information about the current Figma document and user selections, providing context for AI-driven code generation. - Element Creation: Supports the creation of design elements such as rectangles, frames, and text nodes within Figma, allowing AI assistants to build or modify designs programmatically. - Text Content Modification: Enables scanning and updating of text nodes, facilitating bulk text content replacement and updates. - Styling Adjustments: Allows setting of fill colors, stroke colors, and corner radii for design elements, providing control over the visual styling of components. - Layout and Organization: Supports moving, resizing, deleting, and cloning of nodes, aiding in the organization and structuring of design elements. - Annotations: Facilitates the creation and management of annotations within Figma files, enhancing collaboration and documentation. - Component and Style Management: Provides access to local and team components, as well as styles, enabling consistent design implementations. - Export Capabilities: Allows exporting of nodes as images in various formats, supporting the generation of assets from designs. - Code Execution: Supports execution of arbitrary JavaScript code within Figma, enabling advanced customizations and operations. Designed for local operation, the server requires Bun for execution and is actively maintained as an open-source project.

K8M offers a streamlined interface for managing Kubernetes clusters, incorporating AI functionalities such as: - AI-Driven Insights: Utilizes integrated LLMs for tasks like resource explanation, YAML translation, and log analysis. - MCP Integration: Provides visual management of MCP tools, with 49 built-in tools enabling over 100 cluster operations. It can function as an MCP server for other LLM applications. - Multi-Cluster Management: Supports automatic detection and management of multiple clusters using kubeconfig. - Pod Management: Facilitates file operations within pods, real-time log viewing, and shell command execution. - CRD and Helm Support: Automatically discovers and manages Custom Resource Definitions (CRDs) and integrates with Helm for application deployment. These features make K8M a comprehensive tool for efficient and intelligent Kubernetes cluster management.

MCP Proxy facilitates two primary modes of operation: - Stdio to SSE: Acts as a bridge for clients that communicate over stdio (e.g., Claude Desktop) to connect with remote MCP servers using SSE transport. This mode allows clients to interact with SSE-based servers without native SSE support. - SSE to Stdio: Exposes a local stdio MCP server as an SSE server, enabling remote clients to connect via SSE. In this mode, MCP Proxy listens for SSE requests, spawns the local stdio server, and manages the communication between them. These capabilities enhance the flexibility and interoperability of AI assistants and MCP servers across different transport protocols.

MCP Server for Docker allows AI assistants to perform various Docker operations through natural language interactions, including: - Container Composition: Deploy and manage containers using descriptive commands. - Introspection and Debugging: Inspect and debug running containers. - Persistent Data Management: Handle Docker volumes for data persistence. This server is particularly useful for server administrators, tinkerers, and AI enthusiasts aiming to streamline Docker management through AI-driven interfaces.

The Blender MCP Server is a utility server that integrates Blender, the open-source 3D creation suite, with AI assistants through the Model Context Protocol (MCP). This integration enables AI models to directly interact with and control Blender, facilitating prompt-assisted 3D modeling, scene creation, and manipulation. Key features include: - Two-way Communication: Establishes a socket-based connection between AI assistants and Blender, allowing seamless command execution and data retrieval. - Object Manipulation: Enables creation, modification, and deletion of 3D objects within Blender. - Material Control: Allows application and modification of materials and colors on 3D objects.​ - Scene Inspection: Provides detailed information about the current Blender scene, including object properties and scene hierarchy. - Code Execution: Supports running arbitrary Python code within Blender from the AI assistant, enabling advanced operations and customizations. - Asset Integration: Supports integration with external asset libraries such as Poly Haven and Hyper3D Rodin, allowing for enhanced asset creation and import functionalities. Designed for local operation, the server ensures data privacy and security. It is actively maintained as an open-source project.

JetBrains MCP Proxy Server facilitates seamless communication between AI assistants and JetBrains IDEs. It acts as a proxy, forwarding requests from clients to the IDE, enabling AI-driven code assistance and project navigation.

MCP Send Email allows AI assistants to compose and send emails directly using the Resend API. Key features include: - Plain Text and HTML Emails: Supports sending both plain text and HTML-formatted emails. - Scheduling: Allows scheduling emails for future delivery. - CC and BCC Support: Enables adding CC and BCC recipients. - Reply-To Configuration: Allows setting custom reply-to addresses. - Customizable Sender Email: Supports customizable sender email addresses, requiring prior verification with Resend. This integration streamlines the process of sending emails directly from AI interfaces like Cursor or Claude Desktop, eliminating the need to copy and paste email content.

Cloudflare Workers MCP provides both CLI tooling and in-Worker logic to connect AI assistants, like Claude Desktop, to Cloudflare Workers. It allows developers to expose functions or APIs from their applications or services within Cloudflare's developer platform to large language models (LLMs) in coding agents. The package translates TypeScript methods of a Worker into MCP tools that a local Node.js server can expose to MCP clients.

The Apple MCP Server is an open-source utility server that integrates macOS native applications with AI assistants through the Model Context Protocol (MCP). This integration enables AI models to interact seamlessly with Apple's ecosystem, facilitating tasks such as messaging, note management, contact searches, email handling, reminders, calendar events, web searches, and map navigation. Key Features: - Messages: Send and read messages using the Apple Messages app. - Notes: List, search, and read notes from the Apple Notes app. - Contacts: Search contacts for sending messages or retrieving information. - Emails: Send emails with multiple recipients and attachments, search emails with custom queries, schedule emails for future delivery, and manage scheduled emails. - Reminders: List, search, create, and manage reminders with due dates and notes. - Calendar: Search, list, create, and manage calendar events with details like title, location, and notes. - Web Search: Perform web searches using DuckDuckGo and retrieve content from search results. - Maps: Search for locations, save favorites, get directions, drop pins, create and list guides, and add places to guides. Designed for self-hosted operation, the server requires Bun for execution. It is actively maintained as an open-source project.