Stop Searching. Start Trusting.

The curated directory of MCP servers, vetted for security, efficiency, and quality.

Tired of the MCP "Marketplace" Chaos?

We built MCPScout.ai to solve the ecosystems biggest pain points.

No Insecure Dumps

We manually analyze every server for basic security flaws.

Easy Setup

Our gotcha notes warn you about complex setups.

Avoid "Token Hogs"

We estimate token costs for cost-effective agents.

Products, Not Demos

We filter out "Hello World" demos.

CATEGORIES:
SORT:

Vetted Servers(8554)

0
0
Medium Cost
Tochi-Nwachukwu icon

Monopoly-Game-MCP

by Tochi-Nwachukwu

Sec8

This project provides an MCP (Multi-Modal Controller Protocol) server for a Monopoly game, allowing AI agents (OpenAI GPT, Ollama Llama) to compete by making decisions via exposed game tools.

Setup Requirements

  • ⚠️Requires OpenAI API Key (Paid service) for OpenAI agents
  • ⚠️Requires a local Ollama server running and the Llama 3.1 model pulled for Ollama agents
  • ⚠️Python 3.12+ is required
Verified SafeView Analysis
The project avoids direct use of `eval` or other highly dangerous functions. API keys are managed via environment variables. The MCP server exposes specific, well-defined tools, mitigating arbitrary command execution. The `_parse_response` function in `ai_agents.py` attempts to extract JSON from LLM output, including a regex fallback, but it's for parsing tool calls, not executing arbitrary code. External dependencies (OpenAI/Ollama APIs) introduce inherent risks from those services.
Updated: 2025-11-27GitHub
0
0
Medium Cost
moinugare19 icon

ssh-client-mcp

by moinugare19

Sec7

A Model Context Protocol (MCP) server enabling AI assistants like Claude to interact with remote SSH/SFTP servers for command execution and file operations.

Setup Requirements

  • ⚠️Requires Node.js runtime (>=18.0.0).
  • ⚠️Requires proper SSH access and authentication setup (password or private key) for target remote servers.
  • ⚠️Integration is specifically mentioned with Claude Desktop, requiring manual configuration of its 'claude_desktop_config.json' file.
  • ⚠️The local server process needs file system access on the machine where it runs for 'sftp_upload' and 'sftp_download' operations ('localPath').
Verified SafeView Analysis
The server provides direct remote command execution and file transfer capabilities via SSH/SFTP, which inherently grants significant access to target machines. The implementation uses the robust 'ssh2' library and does not contain 'eval' or intentional obfuscation. Sudo password handling is designed to prevent logging, but passing it as part of the command string can lead to brief exposure in process lists on the remote server. Credentials (password or private key path) are loaded from configuration files, environment variables, or CLI arguments; users are responsible for securing these sources. While the 'ConfigManager' has a 'saveConfig' method that, as written, would include passwords if called, the main 'index.ts' entry point does not expose configuration modification tools to the AI, mitigating this risk for typical tool usage. The primary security consideration lies in the powerful access granted to the AI agent and careful management of SSH credentials by the user.
Updated: 2026-01-19GitHub
0
0
Medium Cost
La-Palma-24 icon

lapalma24-mcp

by La-Palma-24

Sec7

Provides a Model Context Protocol (MCP) server for searching and discovering vacation rental properties in La Palma, Canary Islands.

Setup Requirements

  • ⚠️The `API_KEY` environment variable should be set in production to override the hardcoded 'demo_key_12345' default, ensuring appropriate backend access permissions.
  • ⚠️Requires Node.js version 18.0.0 or higher.
  • ⚠️Requires `npm install` to install dependencies and `tsc` to build TypeScript files into JavaScript.
Verified SafeView Analysis
The server uses an API key (`API_KEY`) to access its backend API (`API_BASE_URL`). A default key, 'demo_key_12345', is hardcoded if the environment variable is not set. While the README implies public read-only access for the MCP server itself, the security posture of the backend API with this 'demo_key_12345' is critical. If this demo key grants anything beyond strictly public read-only access to the backend API, or if it's not overridden in a production deployment, it could pose a security risk. No obvious direct code execution vulnerabilities like 'eval' or arbitrary command execution were found.
Updated: 2026-01-18GitHub
0
0
Medium Cost
nlvcodes icon

mcp-server

by nlvcodes

Sec9

Payload CMS Website Template for building blogs, portfolios, and content publishing platforms with a backend admin panel.

Setup Requirements

  • ⚠️Requires environment variables: PAYLOAD_SECRET, DATABASE_URI, NEXT_PUBLIC_SERVER_URL (or VERCEL_PROJECT_PRODUCTION_URL), PREVIEW_SECRET, CRON_SECRET.
  • ⚠️Requires MongoDB instance.
  • ⚠️Initial setup requires creating an admin user via the Payload UI to seed content.
  • ⚠️Uses pnpm for package management.
Verified SafeView Analysis
The server uses environment variables for sensitive configurations like PAYLOAD_SECRET, DATABASE_URI, PREVIEW_SECRET, and CRON_SECRET. Access control is implemented for collections and preview routes, requiring authentication. A hardcoded API key and URL for an AI example exist in a seed content code block, which is for demonstration purposes and not directly executed by the server. Overall, good security practices are followed.
Updated: 2025-11-19GitHub
0
0
Low Cost
itsanamune icon

retellsimp

by itsanamune

Sec9

Provides an MCP interface to the Retell AI API for managing and deploying AI phone and chat agents through compatible clients.

Setup Requirements

  • ⚠️Requires 'RETELL_API_KEY' environment variable to be set.
  • ⚠️Requires Node.js version 18.0.0 or higher.
  • ⚠️A build step ('npm run build') is necessary to compile TypeScript to JavaScript before execution.
  • ⚠️Configuration with Claude clients requires specifying the full path to the compiled 'dist/index.js' file.
Verified SafeView Analysis
The server correctly retrieves the API key from environment variables, avoiding hardcoding. Input validation is performed using JSON schemas for each tool. All external API calls are made to a hardcoded, secure HTTPS endpoint. No dynamic code execution (e.g., eval) or other obvious malicious patterns were found. Error handling for API responses is in place.
Updated: 2025-12-02GitHub
0
0
Medium Cost
Sec8

Provides real-time and forecast weather data, air quality information, and multi-city weather trend analysis via the Model Context Protocol (MCP), integrating with OpenWeather and OpenRouter APIs.

Setup Requirements

  • ⚠️Requires OpenWeather API Key (Paid or Free Tier)
  • ⚠️Requires Node.js and npm installed
  • ⚠️OpenRouter API Key (Paid Tier) is optional for basic analysis, but required for full LLM-powered trend analysis
Verified SafeView Analysis
API keys (OpenWeather, OpenRouter) are correctly handled via environment variables. Input validation is performed using Zod schemas for tool parameters, mitigating common injection risks. Filename sanitization is implemented for saving reports, preventing path traversal. CORS is enabled for all origins, which is explicitly noted in the README as suitable for local development but a security risk for production, requiring proper configuration for deployment. Interaction with OpenRouter (a third-party LLM) introduces an external dependency for analysis, implying data transfer to OpenRouter, but this is clearly part of the server's functionality and not an arbitrary data leak. No 'eval' or obvious malicious code patterns were found.
Updated: 2025-11-20GitHub
0
0
Medium Cost
Sec7

An interactive agentic server for Oracle EPM Cloud Financial Consolidation and Close (FCCS) applications, providing performance dashboards, financial analysis, data reporting, and an RL-enhanced tool execution engine.

Setup Requirements

  • ⚠️Requires access to an Oracle EPM Cloud FCCS instance with appropriate API credentials (FCCS_API_ENDPOINT, FCCS_USERNAME, FCCS_PASSWORD).
  • ⚠️Database initialization is required (python scripts/init_db.py) and defaults to SQLite, but can be configured for PostgreSQL via DATABASE_URL.
  • ⚠️Optional Gmail API credentials (GMAIL_CREDENTIALS_PATH, GMAIL_TOKEN_PATH) are needed for email functionalities.
Verified SafeView Analysis
The server handles sensitive FCCS credentials (username, password, API endpoint) which are loaded from environment variables, a good practice. It exposes a web API locally (http://localhost:8080) and has CORS enabled, which could be a risk if deployed publicly without additional authentication/authorization layers. It also has optional Gmail integration, adding another potential attack surface if misconfigured. The system relies heavily on external calls to the Oracle FCCS API, meaning the security of the FCCS instance itself is paramount. Hardcoded financial data in DCF scripts could be a concern if used for live analysis without dynamic data fetching.
Updated: 2025-12-26GitHub
0
0
Medium Cost
BenjaminDuthe icon

proxmox-mcp

by BenjaminDuthe

Sec6

MCP server enabling Claude to manage Proxmox VE infrastructure for VMs, LXC containers, snapshots, storage, and cluster resources.

Setup Requirements

  • ⚠️Requires a running Proxmox VE instance and API access. Authentication requires either `PROXMOX_TOKEN_ID` and `PROXMOX_TOKEN_SECRET` OR `PROXMOX_USER` and `PROXMOX_PASSWORD`.
  • ⚠️For SSH tools (e.g., `ssh_execute`, `vm_exec_sync`), an SSH key must be generated and authorized on the Proxmox host, and its path explicitly mounted into the Docker container or provided to the local Python environment (e.g., `~/.ssh/id_proxmox`). This involves careful filesystem permissions (`chmod 600`).
  • ⚠️By default, SSL certificate verification for the Proxmox API and SSH host key verification are disabled, which simplifies setup but introduces a security risk (Man-in-the-Middle). For production use, it's recommended to enable verification and configure proper certificates/known hosts.
Review RequiredView Analysis
The server uses `httpx` and `asyncssh` to connect to Proxmox. By default, `PROXMOX_VERIFY_SSL` is set to `false` and SSH `known_hosts` is `None`, which disables certificate and host key verification. While convenient for self-signed Proxmox setups, this creates a man-in-the-middle vulnerability if not addressed with proper certificates. The `vm_file_write` tool has a list of `PROTECTED_PATHS` (e.g., `/etc/shadow`, `/etc/passwd`) preventing writes, but this can be bypassed with `force=True`. Granting an AI access to SSH tools (like `ssh_execute` as root) or file writing inside VMs via guest agent (even with `force=True`) carries significant risk and requires careful permission management of the API token.
Updated: 2026-01-17GitHub
0
0
Low Cost
Sec4

Provides weather and relative date to absolute date conversion utilities as an MCP server for AI models.

Setup Requirements

  • ⚠️Requires an Aliyun Weather API `APPCODE` to fetch weather data. The placeholder in `src/service.ts` must be replaced with a valid key.
  • ⚠️The server listens on port 5000, ensure it's available.
Review RequiredView Analysis
The `src/service.ts` file contains a hardcoded placeholder for `APPCODE` (Aliyun Weather API key). While it's a placeholder, deploying this code as-is would either not function or, if a real key were hardcoded, pose a critical security risk as secrets should be loaded from environment variables (e.g., `process.env`). The `dotenv.config()` call is commented out, indicating an intention to use environment variables that was not fully implemented.
Updated: 2025-11-27GitHub
0
0
High Cost
Sec8

An AI-powered Model Context Protocol (MCP) server designed for intelligent document processing, including text extraction, document type detection, and AI-driven chat with context retention via Google Gemini AI.

Setup Requirements

  • ⚠️Requires a Google Gemini API key, which may involve costs for API usage.
  • ⚠️The application architecture consists of two separate Node.js processes (an 'Express-client' and the 'MCP Server') that need to be run concurrently for full functionality.
  • ⚠️Requires Node.js version 16 or higher.
Verified SafeView Analysis
The server demonstrates good security practices by using environment variables for API keys (`GEMINI_API_KEY`). File uploads are validated for size (10MB limit) and type (pdf, png, jpg, jpeg, webp, txt), mitigating some common attack vectors like large file DoS or unauthorized file execution. The `mcp/package.json` lists `express-rate-limit` and `helmet` as dependencies, indicating an intent for security middleware, though the provided `mcp/index.js` snippet does not explicitly show `helmet` being applied. CORS is enabled in the client, which could be overly permissive if not configured to specific origins in a production environment. No 'eval' or obvious malicious code patterns were found. The use of `@modelcontextprotocol/sdk` implies a structured and potentially more secure communication protocol.
Updated: 2025-12-03GitHub
0
0
Medium Cost
dcstang icon

mcp-trillium

by dcstang

Sec9

Manages and automates operations within a Trilium Notes instance, enabling programmatic search, content manipulation, and attribute updates.

Setup Requirements

  • ⚠️Requires a running Trilium Notes instance accessible at http://localhost:37840/etapi
  • ⚠️Requires Python 3.12 or newer
  • ⚠️TRILIUM_API_KEY environment variable must be set
Verified SafeView Analysis
The server loads API keys from environment variables, avoiding hardcoded secrets. It primarily interacts with a local Trilium Notes API (localhost:37840), limiting direct external network exposure. Markdown to HTML conversion uses standard libraries, and no 'eval' or obvious malicious patterns were found. A minor potential risk is if the Trilium application itself does not fully sanitize HTML content received via its API, which could lead to XSS within Trilium's rendering, but this is a dependency's risk rather than a direct vulnerability in the provided server code.
Updated: 2025-12-11GitHub
0
0
Medium Cost
kinescoper icon

kinescope-mcp

by kinescoper

Sec8

Provides a Model Context Protocol (MCP) server to access Kinescope video platform APIs, enabling AI models to interact with video management, analytics, and live streaming functionalities.

Setup Requirements

  • ⚠️Requires a Kinescope API key (KINESCOPE_API_KEY environment variable or --api-key flag).
  • ⚠️Cursor IDE has an 80-tool limit, which requires careful configuration (`--include-all-tools` with `--resource` filtering or relying on dynamic tools) to avoid exceeding it.
  • ⚠️Requires Node.js version 18.0.0 or higher.
Verified SafeView Analysis
The server uses `JSON.parse` in `src/compat.ts` for handling client-specific JSON serialization quirks (e.g., Claude Desktop). While `zod` validation is applied subsequently, parsing untrusted string input can be a vector for certain types of attacks if not handled with extreme care. However, the context suggests this is to re-parse arguments from specific LLM clients which wrap JSON as strings, and these arguments are then validated by `zod` schemas. No hardcoded secrets were found apart from test API keys in test files. API keys should be handled securely via environment variables or CLI arguments.
Updated: 2025-12-23GitHub
PreviousPage 571 of 713Next