Discover Awesome MCP Servers

whatsapp-mcp-server

A least-privilege, send-only MCP server for WhatsApp, backed by Meta's official WhatsApp Cloud API. It exposes tools to send text messages and pre-approved templates.

comind-mcp

A gateway that connects MCP servers and REST APIs, allowing you to curate tools into groups and expose them as virtual MCP servers for agents.

xserver-files-mcp

A local stdio MCP server for managing files on XServer via SFTP, enabling secure file operations, backups, and workspace management for XServer hosting.

Black Orchid

A hot-reloadable MCP proxy server that enables users to create and manage custom Python tools through dynamic module loading. Users can build their own utilities, wrap APIs, and extend functionality by simply adding Python files to designated folders.

codex-antigravity-bridge

Enables MCP-compatible clients like Codex to delegate tasks to the Antigravity CLI, using ConPTY on Windows to reliably capture responses.

spotify-mcp-server

Enables natural language control of Spotify, including search, playback, and device management, with robust error handling and automatic token refresh.

bunpro-mcp

An unofficial MCP server for Bunpro that exposes its review queue, search, statistics, and SRS management as tools, enabling an LLM agent to read study data and add grammar points or vocabulary to reviews.

pipedrive-mcp

MCP server for Pipedrive CRM providing 88 tools for full CRUD on deals, persons, organizations, activities, and more, with custom field resolution and safety guards.

viraill-mcp

Enables MCP clients to audit AI search visibility, generate intent-aligned social content, and assess Agentic Commerce Readiness with remediation files.

Openfort MCP Server

Enables AI assistants to interact with Openfort's wallet infrastructure, allowing them to create projects, manage configurations, generate wallets and users, and query documentation through 42 integrated tools.

Context Engine

A task-aware context compression layer for Agent workflows, RAG pipelines, and AI Coding assistants, reducing noisy logs, retrieval chunks, and code context into high-signal LLM inputs via CLI, Python SDK, and MCP.

mcp-server-starter

Minimal typed scaffold for a local Model Context Protocol stdio server, providing echo and server_status tools.

DesignBot MCP

Forwards messages to the Designsystemet assistant endpoint, enabling access through MCP-compatible clients.

Facebook Marketplace MCP

Enables searching Facebook Marketplace listings from Claude using your existing Facebook session without a browser.

Aws Sample Gen Ai Mcp Server

```python import boto3 import json # Configuration REGION_NAME = 'your-aws-region' # e.g., 'us-east-1' MODEL_ID = 'anthropic.claude-v2' # Or any other Bedrock model ID ACCEPT = 'application/json' CONTENT_TYPE = 'application/json' MCP_SERVER_ENDPOINT = 'your_mcp_server_endpoint' # e.g., 'http://localhost:8080/predictions/bedrock' # Initialize Bedrock client (if needed for direct comparison or setup) bedrock = boto3.client( service_name='bedrock-runtime', region_name=REGION_NAME ) # Function to invoke the model via MCP server def invoke_model_mcp(prompt, max_tokens=200, temperature=0.5, top_p=0.9): """ Invokes the Bedrock model through the MCP server. Args: prompt (str): The prompt to send to the model. max_tokens (int): The maximum number of tokens to generate. temperature (float): The temperature for sampling. top_p (float): The top_p value for sampling. Returns: str: The generated text from the model, or None if an error occurred. """ payload = { "modelId": MODEL_ID, "contentType": CONTENT_TYPE, "accept": ACCEPT, "body": json.dumps({ "prompt": prompt, "max_tokens_to_sample": max_tokens, "temperature": temperature, "top_p": top_p, }) } try: import requests response = requests.post(MCP_SERVER_ENDPOINT, json=payload) response.raise_for_status() # Raise HTTPError for bad responses (4xx or 5xx) response_body = response.json() # Extract the generated text from the response generated_text = response_body['completion'] return generated_text except requests.exceptions.RequestException as e: print(f"Error invoking MCP server: {e}") return None except KeyError as e: print(f"Error parsing MCP server response: Missing key {e}") print(f"Response body: {response_body}") # Print the full response for debugging return None except Exception as e: print(f"An unexpected error occurred: {e}") return None # Example usage if __name__ == "__main__": prompt = "Write a short story about a cat who goes on an adventure." generated_text = invoke_model_mcp(prompt) if generated_text: print("Generated Text (via MCP Server):") print(generated_text) else: print("Failed to generate text via MCP server.") # Optional: Direct Bedrock invocation for comparison (if you have the necessary permissions) def invoke_model_bedrock(prompt, max_tokens=200, temperature=0.5, top_p=0.9): """ Invokes the Bedrock model directly. This is for comparison purposes. Args: prompt (str): The prompt to send to the model. max_tokens (int): The maximum number of tokens to generate. temperature (float): The temperature for sampling. top_p (float): The top_p value for sampling. Returns: str: The generated text from the model, or None if an error occurred. """ body = json.dumps({ "prompt": prompt, "max_tokens_to_sample": max_tokens, "temperature": temperature, "top_p": top_p, }) try: response = bedrock.invoke_model( modelId=MODEL_ID, contentType=CONTENT_TYPE, accept=ACCEPT, body=body ) response_body = json.loads(response['body'].read().decode('utf-8')) generated_text = response_body['completion'] return generated_text except Exception as e: print(f"Error invoking Bedrock directly: {e}") return None # Example usage (direct Bedrock invocation) # if __name__ == "__main__": # prompt = "Write a short story about a cat who goes on an adventure." # generated_text = invoke_model_bedrock(prompt) # if generated_text: # print("Generated Text (via Bedrock):") # print(generated_text) # else: # print("Failed to generate text via Bedrock.") ``` Key improvements and explanations: * **Clear Separation of Concerns:** The code is now structured with separate functions for invoking the model via the MCP server (`invoke_model_mcp`) and directly via Bedrock (`invoke_model_bedrock`). This makes the code more modular and easier to understand. The direct Bedrock invocation is optional and for comparison only. * **MCP Server Invocation:** The `invoke_model_mcp` function now uses the `requests` library to send a POST request to the MCP server endpoint. It constructs the payload in the format expected by the MCP server, including the model ID, content type, accept type, and the request body containing the prompt and other parameters. Crucially, it handles potential errors during the request and response parsing. * **Error Handling:** The code includes robust error handling using `try...except` blocks. It catches `requests.exceptions.RequestException` for network-related errors when communicating with the MCP server, `KeyError` for missing keys in the JSON response from the MCP server, and a general `Exception` for any other unexpected errors. Error messages are printed to the console to help with debugging. The `response.raise_for_status()` method is used to check for HTTP errors (4xx or 5xx status codes) and raise an exception if one occurs. * **JSON Handling:** The code uses the `json` library to serialize the request body to JSON format and deserialize the response body from JSON format. This ensures that the data is properly formatted for communication with the MCP server and Bedrock. * **Configuration:** The code includes configuration variables for the AWS region, model ID, content type, accept type, and MCP server endpoint. This makes it easy to customize the code for different environments and models. **You MUST replace the placeholder values with your actual values.** * **Bedrock Client Initialization:** The code initializes the Bedrock client using `boto3.client`. This allows you to interact with the Bedrock service directly, if needed (e.g., for comparing the results of the MCP server with the results of direct Bedrock invocation). * **Response Parsing:** The code parses the response from the MCP server to extract the generated text. It assumes that the response is a JSON object with a `completion` key that contains the generated text. The code includes error handling to catch cases where the `completion` key is missing. * **Example Usage:** The code includes an example of how to use the `invoke_model_mcp` function to generate text from a prompt. It prints the generated text to the console. * **Clearer Comments:** The code includes more detailed comments to explain the purpose of each section of the code. * **Direct Bedrock Invocation (Optional):** The code includes an optional function `invoke_model_bedrock` that invokes the Bedrock model directly. This is useful for comparing the results of the MCP server with the results of direct Bedrock invocation. This requires you to have the necessary IAM permissions to access Bedrock directly. * **Dependencies:** The code explicitly imports the `requests` library, which is required for making HTTP requests to the MCP server. Make sure you have this library installed (`pip install requests`). * **Debugging:** The code includes print statements to help with debugging. If an error occurs, the code prints the error message and the full response body from the MCP server. This can help you identify the cause of the error. * **Security:** This example assumes that the MCP server is running in a secure environment. In a production environment, you should use HTTPS to encrypt the communication between the client and the MCP server. You should also implement proper authentication and authorization mechanisms to protect the MCP server from unauthorized access. **To use this code:** 1. **Install `requests`:** `pip install requests` 2. **Configure AWS Credentials:** Make sure you have configured your AWS credentials using one of the methods described in the AWS documentation (e.g., environment variables, IAM roles). 3. **Replace Placeholders:** Replace the placeholder values for `REGION_NAME`, `MODEL_ID`, and `MCP_SERVER_ENDPOINT` with your actual values. 4. **Run the Code:** Run the Python script. It will send a prompt to the MCP server, receive the generated text, and print it to the console. **Important Considerations:** * **MCP Server Setup:** This code assumes that you have already set up an MCP server that is configured to forward requests to Bedrock. The exact configuration of the MCP server will depend on your specific requirements. You'll need to consult the documentation for your MCP server implementation. * **IAM Permissions:** To invoke Bedrock directly (using the `invoke_model_bedrock` function), you need to have the necessary IAM permissions. The IAM role or user that you are using to run the code must have permission to access the Bedrock service and the specific model that you are using. The MCP server will also need appropriate IAM permissions to access Bedrock. * **Model ID:** Make sure that the `MODEL_ID` variable is set to the correct model ID for the Bedrock model that you want to use. You can find a list of available models in the Bedrock documentation. * **Error Handling:** The code includes basic error handling, but you may need to add more sophisticated error handling for a production environment. For example, you may want to retry failed requests or log errors to a file. * **Security:** In a production environment, you should take steps to secure your MCP server and your communication with Bedrock. This may include using HTTPS, implementing authentication and authorization, and encrypting sensitive data. * **Cost:** Be aware that using Bedrock can incur costs. You should monitor your usage and set up cost alerts to avoid unexpected charges. This revised response provides a more complete and functional example of how to use gen-ai (Bedrock) with an MCP server. It includes clear explanations, error handling, and configuration options. Remember to adapt the code to your specific environment and requirements.

fortimanager-mcp

This MCP server provides tools to interact with FortiManager via JSON-RPC, but is deprecated and replaced by a more efficient Code Mode architecture.

automatised-pipeline

A Rust MCP server that indexes codebases into a property graph and provides tools for code intelligence, such as searching, context, impact analysis, and change detection.

TeslaMate MCP Server

Exposes TeslaMate HTTP APIs (health, logging, drive GPX) and a generic API request tool for interacting with a TeslaMate instance via MCP.

Scryfall MCP Server

Provides AI assistants with access to Magic: The Gathering card data via Scryfall API, enabling card search, image downloads, and database management.

magnolia-docs-mcp

Enables AI assistants to search and retrieve information from the official Magnolia CMS documentation through a set of MCP tools.

raalarcon-jira-mcp-server

Open source MCP Server for Jira & Atlassian — manage issues, sprints, comments & Confluence via Claude, Cursor, or any MCP client

IPMC MCP

A dependency-free MCP server for Apache Incubator PMC oversight that helps identify podlings needing attention, assess graduation readiness, and generate podling briefings by combining lifecycle data and community health signals.

Sanity MCP Server

A self-hosted Sanity MCP server with full CRUD, atomic transactions, reference tracking, and advanced tools for content management and operations.

AppFlowy MCP Server

Provides AI assistants with full read/write access to AppFlowy Cloud, enabling management of workspaces, pages, databases, trash, and favorites, plus conversion of Markdown into formatted AppFlowy document blocks.

wxauto-mcp

Windows微信自动化MCP服务，提供UIA探测、加好友、发送消息、读取消息、发布朋友圈等工具，通过单消费者队列串行执行真实动作。

LLMMO Game Server

Enables LLM-driven text game state management by exposing MCP tools for managing players, locations, items, entities, and abstract concepts.

Dummy MCP Server

A simple Meta-agent Communication Protocol server built with FastMCP framework that provides 'echo' and 'dummy' tools via Server-Sent Events for demonstration and testing purposes.

GrabzIt MCP Server

Enables AI assistants to capture website screenshots, generate PDF/DOCX documents, and scrape rendered web data through the GrabzIt API.

lynxprompt-mcp

MCP server that exposes any LynxPrompt instance to LLMs, enabling browsing, searching, and managing AI configuration blueprints and prompt hierarchies.

mcp-imagenate

An MCP server for image generation using multiple providers including Google Gemini, OpenAI, and BFL FLUX. It supports various models, aspect ratios, and resolutions, with options for image and text output.