SuperAPI AI Directory
Partner with us中文AI newsFavorites

All AI tools

35,721 results · Page 138 of 993

mcp-k8s-go

AI developer platforms

mcp-k8s-go is an MCP server that connects to Kubernetes, developed by strowk and hosted on GitHub. It allows users to interact with a Kubernetes cluster through the MCP protocol, providing three types of functionality: prompts, resources, and tools. Users can use natural language or clients (such as Claude Desktop) to list contexts, namespaces, nodes, Pods, etc., and perform operations such as getting events, viewing logs, running commands, and creating and modifying resources. The tool supports multiple installation methods: automatic installation via Smithery or mcp-get, using precompiled binaries from npm, npx, downloading GitHub releases, building from source, or using Docker images. Its core features include browsing Kubernetes resources, managing contexts and namespaces, operating Pods and Services, getting events and logs, executing commands in Pods, etc. During configuration, the configuration file path can be specified via the KUBECONFIG environment variable, and command-line options are supported, such as restricting accessible contexts (--allowed-contexts), read-only mode (--readonly), and hiding sensitive information (--mask-secrets). This server is suitable for developers and operations personnel who need to integrate AI assistants with Kubernetes clusters, and can simplify cluster operations and monitoring.

Read more

mcp-mermaid

AI developer platforms

mcp-mermaid is an MCP (Model Context Protocol)-based server used to let AI dynamically generate Mermaid diagrams. It fully supports all of Mermaid's syntax and features, can configure background color and theme, and supports export to base64, svg, mermaid, file, as well as remotely usable svg_url, png_url, and other formats. It also provides Mermaid syntax validation, making it easier for models to output correct graphics over multiple rounds. This tool can be integrated into desktop applications such as Claude, VSCode, Cline, Cherry Studio, and also supports use through HTTP, SSE, and Streamable protocols on platforms such as Alibaba Cloud, ModelScope, glama.ai, and smithery.ai, and provides a Docker image.

Read more

mcp-mongo-server

AI developer platforms

mcp-mongo-server is a Model Context Protocol (MCP) server for MongoDB that lets AI assistants (such as Claude Desktop and Cursor) interact directly with MongoDB databases. It automatically discovers collections, infers document structure, and executes queries, aggregations, and writes through a standard interface, enabling assistants to read, write, and analyze data using natural language. The server provides a read-only mode by default, which blocks all write paths, and features intelligent ObjectId handling, time conversion, progress reporting, and cancellation, with support for both local stdio and remote HTTP transports, and can connect to standalone, replica set, sharded, and Atlas deployments (including TLS).

Read more

mcp-open-library

AI developer platforms

mcp-open-library is a Model Context Protocol (MCP)-based server for interfacing with the Internet Archive's Open Library API, enabling AI assistants to search for book and author information. It provides seven read-only tools covering bibliographic searches by fields such as title, author, subject, publisher, and ISBN, author lookup by name, retrieval of author details via Open Library key, and retrieval of URLs for book covers and author photos. The server returns JSON projections of Open Library responses rather than raw data, and supports pagination and sorting. No installation or build is required; it can be run directly via npx and connected to MCP clients such as Claude Desktop, and can also be installed by name from the official MCP Registry.

Read more

mcp-opennutrition

AI developer platforms

mcp-opennutrition is a Model Context Protocol (MCP)-based server for accessing the OpenNutrition food database. The database contains over 300,000 food entries, nutrition data, and barcode lookup functionality. OpenNutrition integrates authoritative public sources such as USDA, CNF, FRIDA, and AUSNUT, aiming to solve the problem of fragmented and unreliable nutrition data and provide developers and researchers with transparent, comprehensive, and freely accessible nutrition data. This MCP server supports searching for foods by name, paginated browsing of food lists, retrieving detailed nutrition information by food ID, and finding foods using EAN-13 barcodes. The server can run entirely locally, with all data processing and queries performed locally, requiring no external API calls, ensuring privacy and fast responses.

Read more

mcp-playwright

AI developer platforms

Playwright MCP Server is a Model Context Protocol (MCP)-based server that uses Playwright to provide browser automation capabilities for large language models (LLMs), enabling them to interact with web pages in a real browser environment, take screenshots, generate test code, perform web scraping, and execute JavaScript. This tool is suitable for MCP-supporting clients such as Claude Desktop, Cline, Cursor IDE, and VS Code GitHub Copilot. The latest version v1.0.10 introduces device emulation, supporting 143 real device presets (such as iPhone, iPad, Pixel, Galaxy, and desktop browsers), and supports natural language instructions such as "test on iPhone 13." Installation is flexible: it can be installed via npm, mcp-get, Smithery, or the VS Code CLI; it supports standard stdio mode and HTTP mode; browser binaries can be installed automatically. This server is mainly intended for developers, testers, and AI assistant users who need to integrate browser operations into AI workflows, helping to implement scenarios such as automated testing, web scraping, and remote debugging.

Read more

mcp-redis

AI developer platforms

Redis MCP Server is the official natural language interface from Redis, designed specifically for agent applications to efficiently manage and search data in Redis. Through seamless integration with MCP (Model Content Protocol) clients, it enables AI-driven workflows to interact with structured and unstructured data in Redis. Users can ask questions in natural language, such as "store the entire conversation in a stream," "cache this item," "store a session with an expiration time," "index and search this vector," and more. The server provides tools for various Redis data structures such as strings, hashes, lists, sets, sorted sets, pub/sub, streams, JSON, and more, as well as features such as document search, vector indexing and search, and server management. It supports stdio transport, with streamable-http support planned for the future. Installation methods include installation from PyPI (recommended), installation from GitHub, development installation, and deployment using Docker. Configuration can be done through command-line arguments, environment variables, or EntraID authentication (suitable for Azure Managed Redis). The server supports integrations such as OpenAI Agents SDK, Claude Desktop, VS Code with GitHub Copilot, Augment, and more, and is suitable for scenarios such as AI assistants, chatbots, data search and analysis, and event processing. The project uses the MIT license.

Read more

mcp-searxng

AI developer platforms

mcp-searxng is an MCP server that connects the SearXNG search API to AI assistants, providing privacy-focused web search capabilities for MCP-enabled clients such as Claude and Cursor. It requires users to prepare their own SearXNG instance with JSON search enabled, and supports pagination, filtering, direct answers, HTML/structured text/PDF reading, instance capability discovery and query suggestions, replica failover, caching, proxying, and other features. It can run via local STDIO or Streamable HTTP, and is protected using a static bearer or optional OAuth.

Read more

mcp-server-bigquery

AI developer platforms

BigQuery MCP server is a Model Context Protocol-based server that provides access to Google BigQuery. It enables large language models (LLMs) to inspect database structures and execute queries. The server implements three tools: execute-query (executes SQL queries using BigQuery dialect), list-tables (lists all tables in the BigQuery database), and describe-table (describes the schema of a specific table). The server can be configured via command-line arguments or environment variables. It supports automatic installation into Claude Desktop via Smithery, and also supports manually adding the MCP server configuration in Cursor.

Read more

mcp-server-dumplingai

AI developer platforms

mcp-server-dumplingai is the Model Context Protocol (MCP) server implementation for Dumpling AI, used to integrate Dumpling AI's data scraping, content processing, knowledge management, AI agents, and code execution capabilities into MCP-enabled clients. It covers all Dumpling AI API endpoints, providing data APIs such as YouTube transcription, search, maps, news, and reviews, web scraping, crawling, screenshots, and structured data extraction, document conversion and multi-format content extraction, as well as AI agent completions, knowledge base management, and image generation capabilities. It also supports running JavaScript and Python code in a secure environment. The server features automatic error handling and detailed response formatting, can be installed via Smithery, npx, or manually, and supports configuration and use in clients such as Cursor.

Read more

mcp-server-openai

AI developer platforms

mcp-server-openai is an open-source server based on the MCP protocol, used to let Claude directly call OpenAI models for queries. After users configure this server on the Claude desktop client, they can send requests to OpenAI models and obtain returned results without switching interfaces. The material only provides the repository title and configuration entry point, and does not explain the specific OpenAI models supported, installation commands, dependency requirements, fees, or usage results; the relevant details are not yet provided.

Read more

mcp-server-tavily

AI developer platforms

mcp-server-tavily is an MCP server project based on the Tavily API, used to execute search queries and return results in text format. This project was developed by developer Tomatio13 before the official version was released, supports installation to Claude Desktop via Smithery, can also be configured for use in Cursor, or run locally via Docker Compose. The server implements the search tool, which accepts the required parameter query and the optional parameter search_depth (basic or advanced), and the returned content includes the AI response, URI, and search result titles.

Read more

mcp-server-webcrawl

AI developer platforms

mcp-server-webcrawl is an MCP server for web crawler data and archives, enabling AI clients to filter and analyze web page content according to user instructions or autonomously. It provides a full-text search interface supporting Boolean expressions, and can filter resources by conditions such as resource type and HTTP status. The tool is compatible with a variety of web crawlers, is free and open source, requires Claude Desktop and Python (>=3.10), and is installed via pip install on the command line. In addition to being used by LLMs, it can also serve as an interactive terminal search tool to directly search crawled archives on local or remote hosts.

Read more

mcp-server-youtube-transcript

AI video

mcp-server-youtube-transcript is a Model Context Protocol (MCP) server for directly fetching subtitles and transcripts of YouTube videos. It provides the get_transcript tool, which accepts a YouTube video link, Shorts link, or video ID, and supports specifying the language, outputting timestamps, and filtering ads and sponsored content. It can also optionally enable the analyze_video tool, which uses the TwelveLabs Pegasus video understanding model to summarize or answer questions about videos, requiring configuration of a TwelveLabs API key. This project can be used in MCP-compatible clients such as Claude Desktop via Smithery or manual configuration.

Read more

mcp-snowflake-server

AI developer platforms

mcp-snowflake-server is a Model Context Protocol (MCP)-based server implementation for interacting with a Snowflake database. It supports executing SQL queries through tools and exposes data insights and schema context in the form of resources. Main features include: read_query executes SELECT queries; write_query (enabled with --allow-write) executes INSERT, UPDATE, DELETE; create_table (enabled with --allow-write) creates tables; list_databases, list_schemas, list_tables list databases, schemas, tables; describe_table views table column information; append_insight adds data insights to the memo://insights resource. Resources include memo://insights (a continuously updated memo of data insights) and context://table/{table_name} (if prefetching is enabled, provides a schema summary for each table). It supports installation via Smithery or UVX, and can also be installed locally, requiring configuration of Snowflake credentials and claude_desktop_config.json. Write operations are disabled by default and can be enabled via --allow-write. It supports filtering databases, schemas, or tables via exclusion patterns.

Read more

mcp-tinybird

AI developer platforms

mcp-tinybird is an MCP (Model Context Protocol) server project that Tinybird once published on GitHub, used to connect the Tinybird platform with AI tools that support MCP. According to the repository page information, this repository has been deprecated and is no longer maintained, and the official recommendation is to migrate to the new official implementation, Tinybird MCP Server. Apart from the deprecation notice and migration prompt, the repository page does not provide detailed information such as the project's specific features, installation method, usage steps, applicable scenarios, or limitations.

Read more

mcp-unity

AI developer platforms

mcp-unity is a Model Context Protocol (MCP) implementation for the Unity Editor, used to let AI assistants interact with Unity projects. This package provides a bridge between Unity and a Node.js server that implements the MCP protocol, enabling AI agents such as Cursor, Windsurf, Claude Code, Codex CLI, GitHub Copilot, Google Antigravity, and OpenCode to perform operations in the Unity Editor. It supports manipulating and querying Unity scenes and GameObjects through MCP tools, and can automatically add the Unity Library/PackedCache folder for VSCode-like IDEs to enhance code intelligence.

Read more

mcp-victoriametrics

AI developer platforms

mcp-victoriametrics is a Model Context Protocol (MCP) server implementation for VictoriaMetrics, used to let AI clients integrate with VictoriaMetrics instances and their APIs and documentation. It provides interfaces related to monitoring, observability, and debugging, and supports using almost all of VictoriaMetrics' read-only APIs through natural language, such as querying metrics, listing metrics, analyzing alerting rules, debugging relabeling rules, and more. The project also includes built-in up-to-date documentation, supports offline search, and can integrate with VictoriaMetrics Cloud. It can be tried through a public instance, and also supports binary, Docker, or Helm installation.

Read more

mcp-youtube

AI marketing

mcp-youtube is a Model Context Protocol (MCP) server for YouTube, hosted on GitHub. It uses yt-dlp to download subtitles from YouTube and connects to claude.ai via the Model Context Protocol, enabling Claude to obtain video subtitle content and perform processing such as summarization; for example, a user can ask Claude to "summarize this YouTube video." Before use, yt-dlp must be installed locally (via Homebrew or WinGet), and then the server must be installed via mcp-installer under the name @anaisbetts/mcp-youtube.

Read more