Packages

ScreenHand is an open-source MCP server (stdio transport) that gives AI agents local control over macOS and Windows desktop UIs (via Accessibility APIs) and optionally browser automation (Chrome DevTools Protocol). It exposes a large set of tools for UI inspection, interaction (click/type/keys/scroll/menus/drag), perception/OCR fallbacks, job orchestration, and per-app “app mastery map” learning.

vMLX is a local inference server for Apple Silicon that runs MLX-based text/vision models and provides an OpenAI/Anthropic/Ollama-compatible HTTP API (plus image generation/editing and audio STT/TTS). It also advertises MCP support via a Python dependency.

wati-mcp-server is a Python Model Context Protocol (MCP) server that integrates with the WATI WhatsApp Business API, exposing MCP tools for messaging, contacts, template/broadcast messaging, media file handling, and conversation history retrieval (with pagination/filtering), plus a demo get_weather tool.

ZotLink is a Python MCP server that connects to a local Zotero instance and saves open scholarly preprints (e.g., arXiv and selected open preprint servers) into Zotero, including metadata extraction and attaching PDFs (or validated link fallbacks). It uses browser automation (Playwright) when needed and provides MCP tools for checking Zotero status, listing collections, extracting metadata, and saving papers.

A Python MCP server that searches for flight offers (one-way, round-trip, multi-city) and retrieves offer details using the Duffel flight search API. It exposes MCP tools like search_flights and get_offer_details and requires a Duffel API key provided via environment variable.

just-prompt is an MCP (Model Control Protocol) server that exposes a unified interface to multiple LLM providers (OpenAI, Anthropic, Google Gemini, Groq, DeepSeek, and Ollama). It provides MCP tools to send prompts to one or more models, run prompts loaded from files (optionally writing outputs to disk), list providers/models, and run a multi-model “board” workflow where a “CEO” model decides based on board member responses.

Local AI inference server supporting text generation (LLM), image generation, speech-to-text, and text-to-speech across CPU, GPU (Vulkan/ROCm), NPU (XDNA2), and Apple Silicon. Exposes an OpenAI-compatible REST API on localhost:8000 for drop-in integration with existing tools.

Converts any OpenAPI V3 REST API specification into a working MCP server automatically, enabling AI agents to call any REST API through MCP without writing custom integration code.

Provides an MCP server that lets an agent generate PlantUML diagrams (e.g., sequence/class/C4) and return embeddable URLs or images, plus encode/decode PlantUML for URL sharing.

MCP server implementation for QuickBooks Online integration

A universal preprocessing platform that converts documentation websites, GitHub repositories, and PDFs into structured AI knowledge assets, exportable to 16+ formats including Claude Skills, Gemini Skills, OpenAI GPTs, and vector database formats.

Provides an unofficial MCP server exposing vnstock (Vietnam stock/finance data) functionality as MCP tools, with multiple transport options (stdio, SSE, and HTTP streaming).

Valve's Steam Web API providing access to player profiles, game libraries, achievement data, game news, store information, and community features for the Steam gaming platform.

A blazingly fast, AI-friendly directory tree visualization tool written in Rust. Smart Tree goes beyond traditional tree commands by providing intelligent filtering, multiple output formats, and optimization for both human and AI consumption.

An MCP server that lets LLMs interact with local Anki flashcards by calling AnkiConnect-style operations (decks, note types, CRUD for notes, search, and related metadata) via MCP tools and resources.

Borsa MCP is a FastMCP-based remote/local MCP server that lets LLM agents query unified market and finance data for: BIST (stocks/indices), US stocks and indices (via Yahoo Finance), TEFAS funds, BtcTurk+Coinbase crypto markets, FX/commodities data, macro/TCMB data, KAP news, and provides helper tools like screeners/scanners and regulations. It exposes MCP tools to an agent through a server URL (remote) or via local installation/command.

claude-in-mobile is an MCP server (and also a standalone CLI) that lets an AI client control devices and apps for automation: Android via ADB, iOS Simulator via simctl and optional WebDriverAgent for UI inspection/element tapping, Desktop via a Compose Multiplatform companion, and Aurora OS via audb (with additional on-device Python for gestures). It exposes “tools” such as device listing, screenshots (including annotated screenshots), UI interactions (tap/swipe/input), app control, permissions management, and log/system information.

Code-Mode provides a library/client that lets AI agents execute tool workflows by running TypeScript code in a sandboxed VM and invoking tools registered from UTCP/MCP/HTTP/File/CLI sources. It also offers dynamic tool discovery and auto-generated TypeScript interfaces for IDE/agent guidance.

AI-powered browser testing MCP server that automates end-to-end testing by directing an AI agent to navigate and test web applications. Agents receive natural language test descriptions and return pass/fail results with screenshots.

EDDI (Enhanced Dialog Driven Interface) is a Java/Quarkus multi-agent orchestration middleware for conversational AI. It routes conversations to one or more LLM providers, coordinates tool/API calls, and manages stateful, configurable conversation flows via a lifecycle pipeline and versioned JSON behavior rules. It can run as a Docker container (Kubernetes/OpenShift friendly) and includes built-in tools plus support for whitelisted HTTP-call tools.