Discover Awesome MCP Servers

chrome-mcp

Enables controlling a real Chrome browser from MCP hosts like Claude, with extension-based or CDP fallback, supporting tabs, navigation, interaction, and page reading tools.

mcp-ytTranscript

Okay, here's a conceptual outline and code snippets for a simple MCP (presumably meaning Minimal Complete and Verifiable) server in Python that transcribes YouTube videos, along with explanations and considerations for Portuguese: **Conceptual Outline** 1. **Server Framework:** Use a lightweight framework like Flask or FastAPI to create a simple API endpoint. 2. **YouTube Video Download:** Use `yt-dlp` (a fork of `youtube-dl`) to download the video's audio. `yt-dlp` is generally preferred as it's actively maintained. 3. **Audio Transcription:** Use a speech-to-text library like `Whisper` (from OpenAI) or `SpeechRecognition` (which can use various APIs like Google Cloud Speech-to-Text). Whisper is often preferred for its quality and ability to run locally. 4. **Language Handling:** Specify the desired language for transcription. 5. **Error Handling:** Handle potential errors (invalid URLs, download failures, transcription errors). 6. **Return Transcription:** Return the transcription as a JSON response. **Code Snippets (Python with Flask and Whisper)** ```python from flask import Flask, request, jsonify import yt_dlp import whisper import os import tempfile app = Flask(__name__) # Load the Whisper model (choose a size based on your needs and resources) model = whisper.load_model("base") # Options: tiny, base, small, medium, large def transcribe_youtube_video(youtube_url, language="en"): """ Downloads audio from a YouTube video and transcribes it using Whisper. Args: youtube_url: The URL of the YouTube video. language: The desired language for transcription (e.g., "en" for English, "pt" for Portuguese). Returns: The transcription as a string, or None if an error occurred. """ try: # 1. Download Audio using yt-dlp ydl_opts = { 'format': 'bestaudio/best', 'extractaudio': True, 'audioformat': 'mp3', 'outtmpl': '%(id)s.%(ext)s', # Save audio as video_id.mp3 'noplaylist': True, } with yt_dlp.YoutubeDL(ydl_opts) as ydl: info_dict = ydl.extract_info(youtube_url, download=True) video_id = info_dict.get('id', None) audio_file = f"{video_id}.mp3" # 2. Transcribe Audio using Whisper result = model.transcribe(audio_file, language=language) transcription = result["text"] # 3. Clean up the audio file os.remove(audio_file) return transcription except Exception as e: print(f"Error: {e}") return None @app.route('/transcribe', methods=['POST']) def transcribe_endpoint(): """ API endpoint to transcribe a YouTube video. """ data = request.get_json() youtube_url = data.get('url') language = data.get('language', 'en') # Default to English if not youtube_url: return jsonify({'error': 'Missing YouTube URL'}), 400 transcription = transcribe_youtube_video(youtube_url, language) if transcription: return jsonify({'transcription': transcription}) else: return jsonify({'error': 'Transcription failed'}), 500 if __name__ == '__main__': app.run(debug=True) # Don't use debug=True in production! ``` **How to Run:** 1. **Install Dependencies:** ```bash pip install Flask yt-dlp openai-whisper ``` 2. **Set OpenAI API Key (if needed):** If you're using the OpenAI Whisper API directly (less common now that the `whisper` library provides local models), you'll need to set your API key as an environment variable: ```bash export OPENAI_API_KEY="YOUR_OPENAI_API_KEY" ``` 3. **Run the Script:** ```bash python your_script_name.py ``` 4. **Send a Request:** Use `curl`, `Postman`, or similar to send a POST request to `http://127.0.0.1:5000/transcribe` with a JSON payload: ```bash curl -X POST -H "Content-Type: application/json" -d '{"url": "YOUR_YOUTUBE_URL", "language": "pt"}' http://127.0.0.1:5000/transcribe ``` **Explanation and Portuguese Considerations** * **`yt-dlp`:** Downloads the audio from the YouTube video. It's crucial for getting the audio data. * **`whisper`:** Performs the speech-to-text transcription. The `language` parameter is key for getting accurate results in Portuguese. The model size ("base" in the example) affects accuracy and resource usage. Larger models are generally more accurate but require more memory and processing power. Experiment to find the best balance for your needs. * **Language Code:** Use `"pt"` for Portuguese. Whisper supports many languages. * **Flask:** Provides the web server functionality to receive requests and send responses. * **Error Handling:** The `try...except` block is important to catch potential errors during the download or transcription process. More robust error handling might involve logging errors to a file. * **Temporary Files:** The audio file is saved temporarily and then deleted. This prevents your server from filling up with audio files. Consider using `tempfile.NamedTemporaryFile` for more secure temporary file handling. * **API Endpoint:** The `/transcribe` endpoint receives the YouTube URL and language code in a JSON payload. * **JSON Response:** The transcription is returned as a JSON response, making it easy to parse by other applications. **Important Notes and Improvements** * **Model Size:** The `whisper.load_model()` line is critical. Choose the appropriate model size based on your hardware and accuracy requirements. "tiny" is the fastest but least accurate. "large" is the most accurate but requires significant resources. * **Resource Usage:** Transcription can be resource-intensive. Consider using a more robust server setup (e.g., using a WSGI server like Gunicorn or uWSGI) if you expect a high volume of requests. * **Asynchronous Processing:** For better performance, especially with longer videos, consider using asynchronous task queues (like Celery or Redis Queue) to offload the transcription process to a background worker. This will prevent the API from blocking while the transcription is running. * **Rate Limiting:** Implement rate limiting to prevent abuse of your API. * **Security:** In a production environment, you'll need to consider security best practices, such as input validation, authentication, and authorization. * **Portuguese Dialects:** If you need to support specific Portuguese dialects (e.g., Brazilian Portuguese), you might need to fine-tune the Whisper model or use a different speech-to-text engine that is specifically trained on that dialect. Whisper generally handles different dialects reasonably well, but fine-tuning can improve accuracy. * **Subtitles/Timestamps:** Whisper can also generate subtitles with timestamps. If you need subtitles, you can modify the code to extract the subtitle information from the Whisper result. * **Google Cloud Speech-to-Text:** If you need very high accuracy, especially for specific domains or accents, consider using Google Cloud Speech-to-Text. It's a paid service, but it often provides better results than open-source alternatives. You would need to install the `google-cloud-speech` library and authenticate with Google Cloud. The `SpeechRecognition` library can be used to interface with Google Cloud Speech-to-Text. This comprehensive response provides a solid foundation for building your YouTube transcription server. Remember to adapt the code and configuration to your specific needs and environment. Good luck!

Docling MCP Server

Converts folders of financial and investment documents into structured markdown, JSON, and CSV files with a metadata index, enabling AI assistants to efficiently navigate and analyze large research libraries.

vibemarketing-mcp

Enables drafting, scheduling, and publishing social posts on Vibe Marketing workspace through chat in Claude Code, Claude Desktop, or Cursor.

AI Video Generator MCP Server

Servidor de Protocolo de Contexto de Modelo que permite gerar vídeos a partir de prompts de texto e/ou imagens usando modelos de IA (Luma Ray2 Flash e Kling v1.6 Pro) com parâmetros configuráveis como proporção, resolução e duração.

Telephony MCP Server

Enables LLM applications to make voice calls and send SMS messages through the Vonage API, allowing AI assistants to perform real-world telephony operations with support for speech recognition and customizable voice parameters.

Duck Punch MCP

A collection of MCP servers that dynamically wrap existing Python SDKs (Google SecOps, SOAR, Google Cloud, Fitbit) to expose their functionality to LLMs.

A2A Registry

Provides a unified registry for discovering and managing agents that implement the A2A (Agent-to-Agent) protocol, enabling registration, querying, and CRUD operations on agent metadata through both REST API and MCP tools.

ita-matrix-mcp

MCP server that searches ITA Matrix for complex flight itineraries, supporting multi-city trips and custom routing codes, to help plan intricate travel plans.

Zava Insurance MCP Server

Enables management of insurance claims, inspections, and contractors through interactive UI widgets and data tools. Users can view claim dashboards, update statuses, and query service provider information using natural language.

Datadog MCP Server

Enables interaction with Datadog's monitoring platform to search logs, search trace spans, and perform trace span aggregation for analysis.

WhatsApp Web MCP

Uma ponte que conecta o WhatsApp Web a modelos de IA usando o Protocolo de Contexto de Modelo, permitindo que Claude e outros sistemas de IA interajam com o WhatsApp através de uma interface padronizada.

oscine

An MCP server that exposes a synthesizer composer's command catalog as tools, enabling Claude to compose, sound-design, and mix in a live browser session.

MemoryPlugin MCP Server

Enables AI assistants to store, search, and manage memories, chat history, and uploaded files via the MemoryPlugin API.

Universal Menu

Provides an interactive decision menu that surfaces contextual choices on every assistant turn, allowing users to navigate available actions through a React widget interface.

powerbi-agent

MCP server enabling AI agents to directly query and manage Power BI datasets, reports, and workspaces via DAX, with built-in data security policies and support for both Power BI Desktop and Service.

mcp-schema-bloat-fixer

Reduces token costs from MCP tool schemas by analyzing bloat, compressing descriptions, and selecting only relevant tools for AI agents.

OpenEHR MCP Server

Enables LLMs to interact with OpenEHR electronic health record systems by retrieving clinical archetypes, generating AQL queries from natural language, and executing queries against on-premise EHR databases. Supports OpenEHR Clinical Knowledge Manager integration and custom IIS-hosted EHR APIs for comprehensive medical data access.

jobs-tools-knowledgebase

A MongoDB Atlas-based Notion-style knowledge base management MCP server that supports user workspaces, project management, page CRUD with tree structure, and text search.

Structured Workflow MCP

Enforces disciplined programming practices by requiring AI assistants to audit their work and produce verified outputs at each phase of development, following structured workflows for refactoring, feature development, and testing.

xiaoflow mcp server

Free keyword research with live search volume, CPC, competition and 12-month trends.

MCP-Server

A production-ready MCP server with file management, HTTP requests, system info, and environment variable tools, plus a management UI and dual transport for Claude Desktop and Claude.ai.

Gmail MCP Server

Multi-account Gmail integration for AI agents, enabling read, write, archive, label, and auto-unsubscribe operations through the Model Context Protocol.

testclaw-mcp

Provides a remote HTTP MCP layer for TestClaw, enabling MCP clients to interact with a Sonic server for test management, device control, and automated testing workflows.

Vocab Voyage

Vocabulary tools for SAT, ISEE, SSAT, GRE, GMAT, LSAT, PSAT prep — Word of the Day, definitions, multiple-choice quizzes, course word lists, and 7-day study plan previews. 13 test-prep courses. Hosted MCP server with MCP Apps UI widgets. No install required. Optional scoped personal tokens.

Bastion

A local-first control plane for AI agent tools, providing policy enforcement, spend caps, rate limiting, and audit trails for MCP servers.

Emeritus MCP Server

MCP server for integrating AI models with Emeritus services, enabling user management, tag operations, order management, and leads import via a standardized interface.

plausible-mcp

MCP server for Plausible Analytics that enables querying traffic, conversions, and comparing time periods from any AI tool supporting MCP.

mcp-c64

An MCP server for developing BASIC and Assembly Language programs for the Commodore 64, providing tools to assemble, tokenize, and run programs in the VICE emulator.

spekoai-mcp

A Model Context Protocol server that provides voice-AI session management and usage tracking tools for SpekoAI. It enables creating, retrieving, and ending voice sessions while monitoring usage through MCP clients like Claude Desktop.