MCP servers that ship as code — a repository or a package you install and run locally (usually over stdio). They have no hosted endpoint, so we cannot probe or grade them: instead each one shows its repository, install command, license, stars and when it was last updated.
Collected from GitHub repositories and npm, PyPI, Cargo, Go, Maven, NuGet and RubyGems packages tagged as MCP servers, the official MCP Registry, Docker's MCP catalog, ModelScope, Baidu MCP World, Cline and other public directories; one entry per repository. A server that also offers a hosted endpoint is listed under Remote instead. These entries are not counted in our reachability, verification or readiness statistics.
-
DocsRay MCP Server - AI-powered document analysis and question answering with support for 39+ file formats including PDFs, Office documents, images, audio, and video files. Features intelligent document processing, visual analysis, multilingual support, a
npm MIT -
Peekaboo is a macOS CLI & optional MCP server that enables AI agents to capture screenshots of applications, or the entire system, with optional visual question answering through local or remote AI models.
npm MIT ★ 5,489 -
Local browser control and temporal visual evidence for AI agents — an MCP server that drives Chrome, continuously records it, and answers what-happened questions from retained frames, storyboards, filmstrips, and browser events
cargo ★ 3 -
deepeye-mcpgithub.com/favio8/deepeyeupdated this year
DeepEye - 为纯文本大模型提供视觉能力的 MCP Server(图像描述 / OCR / 视觉问答)
pypi MIT -
Multi-backend GGUF vision-language inference engine (moondream3.1 · VideoChat3 · Qwen2.5-VL · MOSS-VL · VideoMAE/ViViT · SmolVLM) with a unified CLI, Web UI and an MCP server for agents.
NOASSERTION ★ 4 -
Your codebase as a world you can fly into. Any repo becomes an explorable 3D knowledge graph, and Claude Code gets the same map through MCP to answer questions without reading every file.
MIT -
mcp-data-analystgithub.com/flashgari/mcp-data-analystupdated this year
Natural-language data analyst: Claude answers questions by calling tools over a real MCP server (including a real subprocess/stdio test), with a read-only SQL safety guard and a FastAPI + Chart.js dashboard. 57 tests against real infrastructure.
MIT