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(8554)

0
0
Low Cost
gianmarcobz icon

todoist-mcp-server

by gianmarcobz

Sec3

This repository provides deployment configuration for a 'todoist-mcp-server' using Docker on the Railway platform.

Setup Requirements

  • ⚠️Docker required
  • ⚠️Requires the actual server code (Dockerfile and application files) which are not provided.
Review RequiredView Analysis
The provided `railway.json` file is a declarative deployment configuration and does not contain executable code for the server's logic. Therefore, a comprehensive security audit of the 'todoist-mcp-server' is impossible with the given source code. Critical files such as the `Dockerfile` and the application's actual source code are missing, which would contain the server's implementation, dependencies, and potential security vulnerabilities like 'eval' usage, network configurations, or hardcoded secrets. The `railway.json` file itself is safe and merely specifies deployment settings, but the safety of the server it describes cannot be determined without access to its full source code.
Updated: 2026-01-04GitHub→
0
0
High Cost
wkoleilat-happytitan icon

mcp-podcast-scraper

by wkoleilat-happytitan

Sec7

An MCP server for scraping, transcribing, and facilitating AI summarization of podcast episodes from YouTube or RSS feeds.

Setup Requirements

  • ⚠️Requires system-wide installation of `yt-dlp` and `ffmpeg` (e.g., via Homebrew on macOS).
  • ⚠️Requires a Deepgram API Key, which is a paid service (though a free tier is available).
  • ⚠️Requires manual configuration of MCP server settings within the Claude Code or Claude Desktop client.
Verified SafeView Analysis
The `rss-parser` is configured with `rejectUnauthorized: false`, which disables SSL certificate verification for RSS feeds. This could potentially expose users to man-in-the-middle attacks if they interact with a malicious RSS feed server. The server also downloads audio content from external URLs (YouTube, RSS audio links), which always carries an inherent risk, although it utilizes established tools like `yt-dlp` to manage this. The Deepgram API key is handled securely via configuration files or environment variables, with clear instructions against committing secrets. Temporary files are stored locally and are cleaned up on server startup.
Updated: 2025-12-25GitHub→
0
0
Medium Cost
domdomegg icon

google-cal-mcp

by domdomegg

Sec9

This server manages Google Calendar events and settings by acting as an OAuth proxy to the Google Calendar API.

Setup Requirements

  • ⚠️Requires manual creation and configuration of Google OAuth credentials (Client ID, Client Secret) through the Google Cloud Console, including enabling the Google Calendar API and setting up a consent screen.
  • ⚠️Requires setting `GOOGLE_CLIENT_ID` and `GOOGLE_CLIENT_SECRET` as environment variables, particularly for HTTP transport, along with `MCP_TRANSPORT=http`.
Verified SafeView Analysis
The server functions as an OAuth proxy, processing sensitive Google OAuth client credentials (client ID/secret) and access tokens in transit. It explicitly states it holds no long-term tokens or state, which is a strong security practice. Access tokens are used as keys in an in-memory cache for validation, but are not persisted. Environment variables are correctly used for credentials. The OAuth flow implementation (wrapping client redirect URIs in state) appears robust against common redirection attacks. No 'eval' or obvious obfuscation detected. The primary security considerations involve proper management of Google OAuth credentials by the user.
Updated: 2026-01-07GitHub→
0
0
Low Cost
SamEricBailey icon

PlayWright-Framework-PoC

by SamEricBailey

Sec2

Provides an easy-to-use, no-code platform for automated UI, API, and performance testing of web applications.

Setup Requirements

  • ⚠️Requires downloading and executing an opaque '.zip' file containing a pre-compiled application, whose source code is not provided in the repository.
  • ⚠️The README.md specifies a system requirement for 'Version 14.x or later' with a malformed URL in its place, which likely refers to Node.js, a prerequisite for Playwright-based projects that contradicts the 'no coding knowledge / double-click' narrative for the application itself.
Review RequiredView Analysis
The provided 'source code' consists only of a README and a single Playwright test file. The README describes a complete 'application' with a GUI that users are instructed to download as a .zip file from a raw GitHub content URL and 'double-click to run.' The actual source code for this runnable application is not provided in the repository context given. Without the full source code for the application that users are told to execute, a comprehensive security audit is impossible. This method of distribution for a purported 'application' without providing its source is a significant security risk, as the content of the executable .zip file cannot be verified.
Updated: 2026-01-19GitHub→
0
0
High Cost
Rigohl icon
Sec2

High-performance web crawling, scraping, and deep web searching, designed to serve as a powerful tool for AI agents via Model Context Protocol (MCP).

Setup Requirements

  • ⚠️Requires Rust toolchain (Rustc, Cargo) for compilation.
  • ⚠️Requires Go and Zig compilers if respective FFI features are enabled during build.
  • ⚠️Requires Chapel compiler (`chpl`), Python 3 (`pylint`, `flake8`), Node.js (`npx eslint`), and Java Development Kit (JDK) with `checkstyle.jar` (assumed to be in PATH) for various `scan_project` and `analizar_proyecto` functionalities.
  • ⚠️May require `HF_MODELS_PATH` environment variable if HuggingFace integration feature is enabled.
  • ⚠️Complex FFI setup may lead to linker issues on some systems (as indicated by `c_wrapper.c`).
Review RequiredView Analysis
CRITICAL VULNERABILITIES: The server executes external commands (e.g., `cargo`, `npx eslint`, `pylint`, `javac`, `chpl`) on user-provided project paths without sandboxing, leading to arbitrary code execution if a malicious project path is supplied. Additionally, it uses `danger_accept_invalid_certs(true)` in `NuclearBypass`, which makes network communications vulnerable to Man-in-the-Middle (MITM) attacks. Extensive use of FFI (Go, Zig) increases the attack surface, and regexes might be vulnerable to ReDoS if applied to untrusted input. Not safe to run without significant sandboxing and code review for certificate validation.
Updated: 2025-12-01GitHub→
0
0
Low Cost
Porta048 icon

AUTOCAD-MCP

by Porta048

Sec8

Control CAD software (AutoCAD, GstarCAD, ZWCAD) with natural language commands from AI assistants.

Setup Requirements

  • ⚠️Requires Windows OS for COM automation.
  • ⚠️Requires Python 3.8+.
  • ⚠️Requires CAD software (AutoCAD, GstarCAD, or ZWCAD) to be installed and licensed.
  • ⚠️Requires `pywin32` library, which is a Windows-specific dependency.
Verified SafeView Analysis
The server uses `win32com.client` for COM automation, which grants significant control over the CAD application. While this is its intended functionality, a compromised AI assistant or misinterpretation of commands could lead to unintended modifications within CAD drawings. The `process_command` tool relies on regex parsing, preventing direct code injection. File saving via `save_drawing` accepts an arbitrary `file_path`, which could potentially be exploited by a malicious input to save files to unintended locations on the system if the server process has elevated write privileges. However, no `eval` or similar dangerous dynamic code execution patterns were found, and it communicates over standard I/O (stdio) rather than exposing network ports.
Updated: 2026-01-17GitHub→
0
0
Low Cost
Mkscode92 icon
Sec5

Manages and monitors a Minecraft Pocket Edition (MCPE) server, providing capabilities to start, stop, and interact via RCON.

Setup Requirements

  • ⚠️Requires a Java Runtime Environment (JRE) to be installed and accessible on the system PATH.
  • ⚠️Requires a Minecraft Pocket Edition server binary (.jar or similar) to be manually downloaded and its path configured in `config.py`.
  • ⚠️Requires RCON to be enabled and configured on the Minecraft server, with matching host, port, and password details set in `config.py`.
Review RequiredView Analysis
The RCON password and other sensitive configuration details are hardcoded directly within `config.py`. This poses a significant security risk if the configuration file is exposed, as it is not loaded from environment variables or a secure secret management system. The use of RCON itself requires careful network configuration (e.g., firewalling) to prevent unauthorized access to the Minecraft server.
Updated: 2026-01-19GitHub→
0
0
Medium Cost
systempromptio icon

systemprompt-core

by systempromptio

Sec9

Provides a foundational platform for building AI agent orchestration systems, including Model Context Protocol (MCP) server and client implementations, to facilitate AI-callable tools and multi-agent management.

Setup Requirements

  • ⚠️Requires Docker to be running for local PostgreSQL development databases if using the local setup wizard.
  • ⚠️The `SYSTEMPROMPT_PROFILE` environment variable must be set to the path of the profile YAML file for the application to load its configuration.
  • ⚠️A `JWT_SECRET` (minimum 32 characters) is mandatory in the secrets configuration for the core API server to start and function securely.
  • ⚠️At least one AI provider API key (e.g., `GEMINI_API_KEY`, `ANTHROPIC_API_KEY`, `OPENAI_API_KEY`) is required for AI-powered features to be available.
  • ⚠️Critical CLI Bug: The `systemprompt cloud config` command, when setting up external database access, replaces the actual database password with `********` in the saved configuration, making direct external database connections impossible without manual correction.
Verified SafeView Analysis
The project emphasizes robust security practices, including strict linting to forbid unsafe code, direct unwrap/panic usage, and `TODO` comments. Secrets (like `JWT_SECRET` and AI API keys) are explicitly managed via environment variables or separate configuration files, never hardcoded in profiles. SQL queries utilize `sqlx` macros for compile-time validation, effectively preventing SQL injection vulnerabilities. Cloud deployment incorporates a secure two-phase provisioning with a shared Docker registry, tenant-scoped image tags, and strict image validation to prevent injection attacks. However, a critical bug exists in the CLI setup process (`crates/entry/cli/src/cloud/tenant_ops/create.rs`) where the external database URL's password is inadvertently replaced with a masked value (`********`) from the API response when enabling external access. This results in a non-functional connection string being stored locally for the user, hindering direct external database access. This is a setup/configuration bug rather than a runtime exploit, but it is a significant usability and management friction point. No `eval` or similar arbitrary code execution patterns were found in the Rust or TypeScript application logic.
Updated: 2026-01-19GitHub→
0
0
Medium Cost
CyprianFusi icon
Sec8

An intelligent agent that integrates Airbnb search capabilities with persistent memory storage using OpenAI's vector stores, acting as a multi-server Model Context Protocol (MCP) client.

Setup Requirements

  • ⚠️Requires OpenAI API Key (Paid)
  • ⚠️Python 3.13 or higher
  • ⚠️Node.js (for Airbnb MCP server via npx)
  • ⚠️uv package manager
Verified SafeView Analysis
The project follows good practices by loading `OPENAI_API_KEY` from an `.env` file and excluding it from version control. No 'eval' or obvious obfuscation is present. The primary security consideration is the reliance on external `npx` commands for the Airbnb server, which introduces a supply chain risk if the `@openbnb/mcp-server-airbnb` npm package were to be compromised. Temporary file handling for memory storage is done securely using `tempfile.NamedTemporaryFile`.
Updated: 2025-12-01GitHub→
0
0
Medium Cost
BACH-AI-Tools icon

mcp-server-chart

by BACH-AI-Tools

Sec7

Generates diverse data visualization charts using the AntV library, serving as a Model Context Protocol server for chart generation and data analysis.

Setup Requirements

  • ⚠️Chart rendering relies on an external AntV API (https://antv-studio.alipay.com/api/gpt-vis) by default; private deployment requires self-hosting 'GPT-Vis-SSR'.
  • ⚠️Geographic map generation tools are limited to regions within China and utilize AMap service.
  • ⚠️Generating records via 'SERVICE_ID' requires obtaining an ID through an Alipay mini-program, which may be a regional friction point.
Verified SafeView Analysis
The server uses Zod for robust input validation. While the server code itself appears secure with no 'eval' or evident hardcoded secrets, its core chart generation relies on an external AntV API ('VIS_REQUEST_SERVER'). The security of the rendered charts and potential data leakage depends heavily on the trustworthiness and security of this external service. The HTTP streamable transport enables CORS for all origins ('*').
Updated: 2025-12-04GitHub→
0
0
Low Cost
Alan-Chen-au icon

mcp-server-demo

by Alan-Chen-au

Sec7

Demonstrates an MCP (Multi-Agent Communication Protocol) server using the FastMCP 2.0 framework, exposing tools, resources, and prompts via an HTTP API.

Setup Requirements

  • ⚠️Requires Python 3.12 or newer
Verified SafeView Analysis
The server enables CORS for all origins (`allow_origins=["*"]`). While this is stated as being for an 'MCP Inspector' and for a 'Demo', it presents a security risk (e.g., CSRF/XSS) if deployed in a production environment without proper upstream security measures. No other critical vulnerabilities like `eval` or hardcoded secrets were found.
Updated: 2025-12-12GitHub→
0
0
Medium Cost
mcpdeploy icon

sample-mcp-server

by mcpdeploy

Sec1

A sample server implementation for the Minecraft Protocol, likely intended for educational purposes, custom game development, or as a base for a private Minecraft server.

Setup Requirements

  • ⚠️Requires a specific runtime environment (e.g., Java, Node.js, Python) which cannot be determined without the source code.
  • ⚠️Network port configuration (e.g., default Minecraft port 25565) will be required for external access and multiplayer connectivity.
Review RequiredView Analysis
No source code was provided for analysis despite the prompt indicating its presence. Without the actual source code, a security audit is impossible. Running any server without inspecting its code is extremely risky, as it could contain arbitrary malicious code, severe vulnerabilities, or hardcoded secrets.
Updated: 2025-11-24GitHub→
PreviousPage 435 of 713Next