Stop Searching. Start Trusting.

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

Browse DirectoryWhy Us?

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(9120)

0
0
Medium Cost
pritamp-cmyk icon

mcp-server

by pritamp-cmyk

Sec4

This server serves as an endpoint for the Model Context Protocol, likely exposing AI models or services for interaction.

Setup Requirements

  • ⚠️Requires Node.js version 18 or higher.
  • ⚠️Relies on the @modelcontextprotocol/sdk, which likely entails external API keys, configuration, or integration with the Model Context Protocol ecosystem.
  • ⚠️The full application logic in `src/index.ts` is not provided, making it impossible to ascertain all configuration requirements.
Review RequiredView Analysis
The main application logic file, `src/index.ts`, is truncated, making a comprehensive security audit impossible. Without this code, it's impossible to check for direct vulnerabilities such as `eval` usage, hardcoded secrets, improper input validation, or other malicious patterns. The project relies on `@modelcontextprotocol/sdk` and its transitive dependencies, which include web server frameworks (`express`, `hono`) and security features like `express-rate-limit` and `jose`. However, the security of the server ultimately depends on the unprovided application logic.
Updated: 2026-01-19GitHub→
0
0
Low Cost
Nordicvitamin icon

docling-mcp-server

by Nordicvitamin

Sec3

A central server potentially for managing or processing documents, acting as a core backend component.

Review RequiredView Analysis
A detailed security audit is impossible as no source code was provided for analysis. Therefore, no checks for 'eval', obfuscation, network risks, hardcoded secrets, or malicious patterns could be performed. Running unvetted code carries inherent risks, and its safety cannot be confirmed.
Updated: 2025-11-19GitHub→
0
0
Low Cost
Knuckles-Team icon

tunnel-manager

by Knuckles-Team

Sec7

This project provides a FastMCP server exposing SSH connection and file transfer capabilities, enabling AI agents to programmatically perform remote server management and automation.

Setup Requirements

  • ⚠️Requires existing SSH configuration (e.g., ~/.ssh/config, private keys) on the host running the server or explicit credentials.
  • ⚠️Remote hosts must be network-accessible via SSH from the server machine (firewall rules, routing, etc.).
  • ⚠️For inventory-based operations, an Ansible-style YAML inventory file is required, specifying host connection details.
Verified SafeView Analysis
The server uses `paramiko.AutoAddPolicy()` which automatically adds unknown host keys, increasing susceptibility to man-in-the-middle attacks if not in a controlled environment. Direct calls to `os.system` for `ssh-keygen` are made, though arguments appear controlled. The Docker setup mounts `~/.ssh` as read-only into the container, granting access to host SSH keys, which is a necessary but high-privilege operation. Password-based authentication can be used for initial setup, which typically involves passing passwords directly.
Updated: 2026-01-19GitHub→
0
0
Medium Cost
jarosser06 icon

mosaic

by jarosser06

Sec8

Personal work memory and time tracking MCP server for managing work sessions, meetings, projects, notes, tasks, and reminders with natural language queries and privacy controls.

Setup Requirements

  • ⚠️Requires PostgreSQL 16.
  • ⚠️Requires the 'uv' package manager for local development.
  • ⚠️The DATABASE_URL environment variable must be configured in a .env file or directly.
  • ⚠️macOS users must sign their Python executable for desktop notifications to function correctly (e.g., `codesign -s - $(which python3)`).
Verified SafeView Analysis
Uses SQLAlchemy ORM to mitigate SQL injection risks. Database credentials are externalized via environment variables. The server is designed as a single-user system, simplifying access control but meaning a compromise exposes all user data. Desktop notification delivery on macOS requires Python executable signing, which can cause silent failures if not configured, but does not pose a direct code vulnerability. The default PostgreSQL password 'changeme' in the Docker Compose example should be updated for any deployment beyond local development.
Updated: 2026-01-17GitHub→
0
0
Medium Cost
mjordi icon
Sec9

Enables AI assistants to search, retrieve, and manage photos, collections, and user profiles from the Unsplash API.

Setup Requirements

  • ⚠️Requires an Unsplash API Access Key (requires account creation and application setup on Unsplash Developers).
  • ⚠️Demo Unsplash API accounts are limited to 50 requests per hour, necessitating application for production access for higher limits.
  • ⚠️Requires Node.js 18 or higher.
Verified SafeView Analysis
The server uses environment variables for the API key (UNSPLASH_ACCESS_KEY), which is good practice. Input validation is handled robustly using Zod schemas, mitigating common injection risks by ensuring parameters conform to expected types and formats. The `unsplash_track_download` tool fetches a download location URL from the trusted Unsplash API response, rather than using arbitrary user input, thereby reducing the risk of Server-Side Request Forgery (SSRF). No usage of `eval`, obfuscation, or other immediately apparent malicious patterns were identified. Error handling for API requests is implemented to catch and propagate Unsplash-specific errors.
Updated: 2026-01-19GitHub→
0
0
Medium Cost
ry-ops icon
Sec9

Integrates Claude AI with Pulseway PSA to automate management of service desk tickets, finance, CRM, and time tracking operations.

Setup Requirements

  • ⚠️Requires Python 3.10 or higher.
  • ⚠️Requires `uv` package manager for dependency installation.
  • ⚠️Requires a Pulseway PSA account with API access and specific API credentials (Gateway URL, Username, Password, Company Name).
  • ⚠️Manual configuration of Claude Desktop's `claude_desktop_config.json` file, including specifying an absolute path to the server directory.
Verified SafeView Analysis
The project strongly emphasizes secure handling of credentials, using environment variables via `secrets.env` (which is git-ignored) and recommending restricted file permissions. It uses standard HTTP(S) communication with basic authentication. There are no signs of `eval`, obfuscation, or malicious patterns. Error handling is present, but a broad `try-except` in `call_tool` could be more specific, though it returns the error to the LLM.
Updated: 2025-12-03GitHub→
0
0
High Cost
shjanjua icon

pptx-mcp-server

by shjanjua

Sec6

A Model Context Protocol (MCP) server that enables AI assistants to create, edit, and manipulate PowerPoint presentations programmatically.

Setup Requirements

  • ⚠️Requires Python 3.10+
  • ⚠️LibreOffice must be installed for thumbnail generation and document validation (e.g., `brew install --cask libreoffice` on macOS, `sudo apt-get install libreoffice` on Linux)
  • ⚠️Poppler-utils (pdftoppm) or ImageMagick (convert) must be installed for thumbnail generation (e.g., `brew install poppler` on macOS, `sudo apt-get install poppler-utils` on Linux)
Verified SafeView Analysis
The server uses `subprocess.run` to invoke external binaries (LibreOffice, poppler-utils/ImageMagick) for thumbnail generation and document validation. While `shell=False` is the default for `subprocess.run` with list arguments, passing user-controlled file paths (`pptx_path`, `office_file`) to these external tools can introduce a risk of command injection if the filenames themselves are crafted with shell metacharacters and the external tools fail to properly sanitize their arguments. Processing untrusted `.pptx` files is a potential vector for this. However, the project wisely uses `defusedxml` for XML parsing, mitigating common XML-related vulnerabilities.
Updated: 2025-12-12GitHub→
0
0
Low Cost
ongzhixian icon
Sec9

Provides an example implementation of an MCP server written in C#.

Setup Requirements

  • ⚠️.NET SDK required
Verified SafeView Analysis
The provided source code consists only of configuration files. No critical vulnerabilities, hardcoded secrets, 'eval' equivalents, or malicious patterns were found in these files. 'AllowedHosts: "*"' is set, which is common for development but should be restricted in a production environment. A full security audit would require the application logic files.
Updated: 2025-12-10GitHub→
0
0
Low Cost

fvtt-mcp-module

by darkbard81

Sec2

A Foundry VTT module that exposes game data and facilitates real-time communication via a WebSocket relay to external services, enhancing interactivity and enabling features like chat logging and audio playback.

Setup Requirements

  • ⚠️Requires Foundry Virtual Tabletop (VTT) installation.
  • ⚠️Requires an external WebSocket relay server (default: wss://foundryvtt-rest-api-relay.fly.dev/).
  • ⚠️Only functions for a 'full GM' (role 4) user who is also the 'primary GM' (lowest user ID among active full GMs).
Review RequiredView Analysis
CRITICAL: The `execute-js` route allows arbitrary JavaScript execution via `eval(script)`. This is a severe remote code execution vulnerability. Additionally, the API key is transmitted as a query parameter in the WebSocket URL, which can expose it in server logs or network sniffers.
Updated: 2025-11-24GitHub→
0
0
Medium Cost
yash-rewalia icon

test-remote-mcpserver

by yash-rewalia

Sec9

Manages personal expense entries, allowing users to add, list, and summarize expenditures.

Setup Requirements

  • ⚠️The `pyproject.toml` specifies `requires-python = ">=3.14"`, which refers to a future Python version not yet released. This is likely a typo and should be adjusted to a currently available version (e.g., `>=3.9` or `>=3.11`) for the project to be installable.
  • ⚠️The `pyproject.toml` lists `aiosqlite` twice and also includes `aisqlite>=1.1.0`. The latter might be a typo for `aiosqlite`, potentially leading to unnecessary dependency installations or conflicts.
Verified SafeView Analysis
The server uses parameterized queries, effectively mitigating SQL injection risks. It explicitly handles database initialization and checks for write access to the temporary database file. No 'eval' or obfuscation is present in the provided code. The server binds to 0.0.0.0, which makes it accessible from external interfaces if deployed without a firewall, but this is a deployment concern rather than a code vulnerability.
Updated: 2025-11-30GitHub→
0
0
Medium Cost
Sec8

A hyper-personalized group chat assistant using real-time contextual data from multiple MCP servers to facilitate group planning.

Setup Requirements

  • ⚠️Requires Anthropic API Key (Paid)
  • ⚠️Requires Google Maps API Key (Paid, requires specific APIs enabled: Directions, Geocoding, Maps JavaScript)
  • ⚠️Requires Weather API Key
Verified SafeView Analysis
The application runs as a web server and interacts with external APIs (Anthropic, Google Maps, Weather). It emphasizes good security practices like storing API keys in a .env file and not committing them to version control. The architecture uses MCP to control tool execution, aiming to reduce the risk of arbitrary code or data access.
Updated: 2025-11-17GitHub→
0
0
Low Cost

datastore-mcp

by hashkrish

Sec7

Provides a Model Context Protocol (MCP) server for AI assistants to interact with Google Cloud Datastore, enabling data management and querying capabilities.

Setup Requirements

  • ⚠️Requires Python 3.10+ or Docker Engine 20.10+ with Docker Compose v2.0+.
  • ⚠️Google Cloud Platform project with Datastore enabled and service account credentials are required for production use.
  • ⚠️Manual startup of the Google Cloud Datastore emulator (`gcloud beta emulators datastore start`) is required for local Python development if not using Docker Compose.
Verified SafeView Analysis
The server provides direct programmatic access to Google Cloud Datastore (or an emulator). While no 'eval' or obvious obfuscation is present in the provided README, direct database manipulation by an AI assistant requires careful configuration of credentials (especially for production GCP) and running within a secure, trusted environment. Misconfiguration or exposure to untrusted AI could lead to data modification or exfiltration.
Updated: 2025-11-17GitHub→
PreviousPage 286 of 760Next