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)

34
29
High Cost
linw1995 icon

nvim-mcp

by linw1995

Sec3

A Model Context Protocol (MCP) server that provides seamless integration with Neovim instances, enabling AI assistants to interact with the editor through connections and access diagnostic information.

Setup Requirements

  • ⚠️Requires Neovim (version 0.11.3+ is mentioned for integration testing) to be installed and accessible in the system's PATH.
  • ⚠️For 'Dynamic Tool System' and optimal auto-connection, the `nvim-mcp` Lua plugin must be installed within Neovim's configuration (e.g., via `lazy.nvim`).
  • ⚠️Many LSP-related tools (e.g., `lsp_code_actions`, `lsp_hover`) require specific Language Servers (e.g., `gopls` for Go, `ts_ls` for TypeScript) to be configured and running within the connected Neovim instance.
Review RequiredView Analysis
The server includes an `exec_lua` tool that allows executing arbitrary Lua code within the connected Neovim instance. While this is an intended feature for powerful integrations, it poses a critical security risk. If the server is exposed via HTTP (`--http-port` with `--http-host 0.0.0.0`) without robust external authentication/authorization, a malicious actor gaining network access to the server could execute arbitrary code within the user's Neovim, potentially compromising the editor and underlying system. There are no explicit hardcoded secrets found in the provided code.
Updated: 2026-01-02GitHub→
34
17
Medium Cost
honeybadger-io icon

honeybadger-mcp-server

by honeybadger-io

Sec9

Provides structured access to the Honeybadger API through the Model Context Protocol (MCP) for integration with LLM agents and MCP clients.

Setup Requirements

  • ⚠️Requires a Honeybadger Personal Auth Token (a personal API key) for authentication with the Honeybadger API.
  • ⚠️Requires Docker to run the pre-built image, or a Go development environment (Go 1.18+ recommended) to build from source.
  • ⚠️This is an MCP server, requiring an MCP-compatible client (e.g., Cursor, VS Code, Zed, Claude Code) to interact with it.
  • ⚠️The server runs in read-only mode by default, meaning write operations (like 'create_project', 'delete_project') are disabled unless 'HONEYBADGER_READ_ONLY=false' is explicitly set.
Verified SafeView Analysis
The server is implemented in Go, which benefits from compile-time safety and reduces certain vulnerabilities common in scripting languages. It operates using STDIO, significantly limiting its direct network attack surface. A critical security feature is its default read-only mode, which prevents destructive API operations (like deleting projects) unless explicitly disabled, enhancing safety against accidental or malicious actions. Authentication relies on an API token passed via environment variables, command-line flags, or a configuration file, avoiding hardcoded secrets. All communication with the Honeybadger API is expected to occur over HTTPS. No 'eval' or similar dynamic code execution patterns, obfuscation, or other malicious patterns were identified within the provided source code.
Updated: 2026-01-05GitHub→
34
4
Low Cost
Nickha124 icon

autumn-mcp

by Nickha124

Sec1

Develop and manage MCP (Minecraft-compatible) servers on the autumn framework.

Setup Requirements

  • ⚠️Software distribution via direct raw GitHub link (.zip) is unusual and may lack integrity verification for executables.
  • ⚠️The Linux installation command `tar -xvzf [URL]` is syntactically incorrect for directly extracting from a remote URL; typically requires downloading the file first (e.g., with `curl` or `wget`).
  • ⚠️Requires active Internet connection for updates (as stated in System Requirements).
Review RequiredView Analysis
Critical: No actual source code was provided for analysis beyond the README.md file. Therefore, a thorough security audit for patterns like 'eval', obfuscation, network risks, hardcoded secrets, or malicious patterns could not be performed. The provided 'SOURCE CODE' section was identical to the README. This score reflects the inability to audit the code, not necessarily confirmed vulnerabilities.
Updated: 2026-01-19GitHub→
34
3
Medium Cost
mimersql icon

mimer-mcp

by mimersql

Sec9

The Mimer MCP Server enables AI applications to securely interact with Mimer SQL databases for schema browsing, read-only query execution, and stored procedure management via the Model Context Protocol.

Setup Requirements

  • ⚠️Requires Mimer SQL 11.0 or later (or the official Mimer SQL Docker container).
  • ⚠️Mandatory database connection environment variables (DB_DSN, DB_USER, DB_PASSWORD, DB_HOST) must be configured in a .env file.
  • ⚠️Running the server in Docker rootless mode might cause network connectivity issues with the database.
Verified SafeView Analysis
The server demonstrates strong security practices for database interaction. The `execute_query` tool explicitly restricts operations to `SELECT` queries, preventing DML/DDL abuse. Both `execute_query` and `execute_stored_procedure` utilize parameterized queries, mitigating SQL injection risks. Configuration for sensitive database credentials (`DB_DSN`, `DB_USER`, `DB_PASSWORD`, `DB_HOST`) is properly handled via environment variables and `.env` files, avoiding hardcoded secrets. While the `entrypoint.sh` script uses `eval`, its usage is limited to controlled environment variable assignments, posing minimal risk. The server listens on all interfaces (`0.0.0.0`) when using HTTP transport, which is typical for containerized applications but should be considered in network security posture.
Updated: 2026-01-12GitHub→
34
4
Medium Cost
quanticsoul4772 icon

analytical-mcp

by quanticsoul4772

Sec9

Provides a suite of analytical tools for data analysis, logical reasoning, and research verification, designed to enhance AI capabilities for structured problem-solving and decision-making.

Setup Requirements

  • ⚠️Requires `EXA_API_KEY` for research functionality (Exa.ai is a paid service).
  • ⚠️Requires Node.js (likely v18+) and npm/yarn to run.
  • ⚠️Persistent caching, if enabled, needs write access to a `./cache` directory.
Verified SafeView Analysis
The server demonstrates robust security practices, including extensive input validation (Zod, custom ValidationHelpers), comprehensive error handling (custom AnalyticalError classes with recovery strategies), and careful management of external API interactions (ResilientApiWrapper, RateLimitManager). API keys (e.g., EXA_API_KEY) are managed via environment variables and checked, preventing hardcoded secrets. File parsing uses SecureFileParser with schema validation and data freezing, reducing data manipulation risks. No direct 'eval' or obvious malicious patterns are present in the truncated code.
Updated: 2026-01-19GitHub→
34
1
Medium Cost
mvilanova icon

tempoai-mcp-server

by mvilanova

Sec7

The Tempo AI MCP Server connects large language models like Claude and ChatGPT with the Tempo AI API to retrieve fitness and wellness data.

Setup Requirements

  • ⚠️Requires Python 3.13 or higher.
  • ⚠️Requires a Tempo AI API Key, which must be obtained from jointempo.ai.
  • ⚠️Initial setup for Claude Desktop or ChatGPT requires specific `mcp install` or environment variable configurations, including potentially using `uv` for package management.
  • ⚠️For ChatGPT, running in SSE mode requires port forwarding (e.g., `ngrok`) for public access.
Verified SafeView Analysis
The server handles API keys by loading them from environment variables or a .env file with appropriate permissions (chmod 600). The `install.sh` script uses `curl | bash` which is a common but sometimes debated practice for initial installation. Runtime execution uses `subprocess.run` via the `mcp` CLI, which is a controlled and expected interaction within the SDK. No direct use of `eval` or obvious obfuscation was found. Communications are over HTTPS to the Tempo AI API.
Updated: 2025-12-03GitHub→
34
3
Medium Cost
gander-tools icon

osm-tagging-schema-mcp

by gander-tools

Sec9

Provides OpenStreetMap (OSM) tagging schema information, validation, and improvement suggestions for AI agents and LLM applications.

Setup Requirements

  • ⚠️Requires Node.js 24.0.0 or higher and npm 11.0.0 or higher.
  • ⚠️Not a standalone application; requires integration with an MCP client (e.g., Claude Code, Claude Desktop) to be useful.
  • ⚠️Initial schema loading can introduce latency on the first call, though preloading is implemented.
Verified SafeView Analysis
The project demonstrates strong security practices including TDD, property-based fuzzing, SLSA Level 3 provenance, SBOM generation, Docker image signing with Cosign, and automated vulnerability scanning (Trivy). It explicitly addresses a known ReDoS vulnerability in the MCP SDK, stating the project is not affected due to not using the vulnerable code paths. CORS is configurable for HTTP transport. No 'eval' or obvious malicious patterns were found. Permission checks for triggering Docker builds in PRs mitigate code injection risks for contributors.
Updated: 2026-01-19GitHub→
34
2
Low Cost
mrorigo icon
Sec9

Orchestrates multiple Model Context Protocol (MCP) servers, enabling LLMs to generate and execute TypeScript code for complex, multi-step workflows, and integrating with Agent-to-Agent (A2A) protocols for collaborative AI.

Setup Requirements

  • ⚠️Requires an API key for a compatible LLM provider (e.g., OpenAI, Anthropic), which are typically paid services.
  • ⚠️Requires Node.js 18.0.0 or higher.
  • ⚠️Integrating with other MCP servers requires those servers to be set up and configured separately.
Verified SafeView Analysis
The project uses a secure sandboxed `vm` environment for executing LLM-generated TypeScript code. Dangerous Node.js globals (`process`, `require`, `__dirname`, `__filename`, `global`, `Buffer`, `module`, `exports`) are explicitly blocked. Execution includes a configurable timeout. All external interactions (e.g., file system, network) must occur through registered MCP tools, which can be controlled and audited. The `SamplingSecurityManager` further enhances security by implementing approval workflows, rate limiting, and policy enforcement for LLM sampling requests, ensuring human-in-the-loop oversight for sensitive operations. No direct `eval` calls in critical paths were found, and code is compiled into a script for execution. The overall design prioritizes secure execution of untrusted code.
Updated: 2025-12-14GitHub→
34
4
High Cost
huseyindeniz icon

gitops-lab

by huseyindeniz

Sec3

A GitOps lab demonstrating multi-cluster application deployment and management (local-management, local-staging, local-production) for AI/ML inference (image background removal), multiplayer gaming, and .NET microservices using Argo CD, Helm, Istio, and related Kubernetes tooling.

Setup Requirements

  • ⚠️Requires a Kubernetes cluster (e.g., Minikube) with Argo CD and Istio as foundational prerequisites for the GitOps setup.
  • ⚠️MetalLB is configured with BGP peering for LoadBalancer services in local clusters, which demands intricate network setup and understanding of BGP for successful operation.
  • ⚠️Optimal performance for AI/ML workloads (such as the image inference API, NVIDIA Triton, and Ollama) is dependent on NVIDIA GPUs for hardware acceleration; otherwise, they will fall back to significantly slower CPU processing.
Review RequiredView Analysis
Critical security risks are present due to hardcoded sensitive credentials (e.g., 'Harbor12345' for Harbor, 'minio123' for MinIO) directly within Helm values files. These are used for critical infrastructure components like the container registry, object storage, and logging storage across both staging and production configurations. Additionally, the metrics server in the 'local-production' environment is configured with '--kubelet-insecure-tls', which is an insecure practice. While the frontend includes Web3 wallet signing, client-side signature verification is noted as a potential weakness if not complemented by robust server-side authentication.
Updated: 2025-12-07GitHub→
34
4
Low Cost
moneyforward-i icon

admina-mcp-server

by moneyforward-i

Sec9

Provides a Micro-service Communication Protocol (MCP) server to interact with the Admina API for managing IT assets, identities, services, and accounts within an organization.

Setup Requirements

  • ⚠️Requires 'ADMINA_API_KEY' environment variable for Admina API authentication.
  • ⚠️Requires 'ADMINA_ORGANIZATION_ID' environment variable to specify the target organization.
  • ⚠️Requires a Node.js runtime environment to execute.
Verified SafeView Analysis
The server demonstrates strong security practices: API keys and organization IDs are correctly handled via environment variables, input validation is rigorously enforced using Zod schemas, and network requests are made via the well-established Axios library to a fixed HTTPS base URL. Error handling is also comprehensive. No 'eval', obfuscation, or hardcoded secrets were found.
Updated: 2026-01-13GitHub→
34
4
Medium Cost
CapulusCodeNinja icon

mcp_nexus

by CapulusCodeNinja

Sec7

Provides AI systems with structured access to Windows debugging tools (WinDBG/CDB) for automated crash dump analysis.

Setup Requirements

  • ⚠️Requires .NET 8.0 Runtime or SDK to be installed.
  • ⚠️Requires Windows Debugging Tools (WinDBG/CDB) to be installed separately.
  • ⚠️Operates exclusively on Windows due to its reliance on WinDBG/CDB.
Verified SafeView Analysis
The server exposes powerful Windows debugging tools (WinDBG/CDB) and allows execution of PowerShell scripts, which can perform file system operations (e.g., downloading source code, managing dump files). If the AI agent or the MCP server itself is compromised, these capabilities could be exploited to execute arbitrary commands or access/modify files on the host system. While IP rate limiting is implemented, the inherent power of the exposed tools constitutes a significant attack surface if not properly secured and isolated in a production environment. No direct 'eval' or hardcoded secrets were identified in the provided code snippets.
Updated: 2026-01-18GitHub→
34
2
High Cost
Lumosylva icon

lottery-mcp-server

by Lumosylva

Sec8

Serves historical Chinese Double Color Ball lottery data and statistical analysis via Model Context Protocol.

Setup Requirements

  • ⚠️Requires Node.js and npm for dependency installation.
  • ⚠️Relies on an external API (www.cwl.gov.cn) for live data, which may change its structure or cookie requirements. The fallback data is a small static subset.
  • ⚠️Needs write permissions in its working directory for 'cookie-cache.json' and 'lottery-data-cache.json'.
Verified SafeView Analysis
The server interacts with an external lottery API (www.cwl.gov.cn) and implements automatic cookie management by fetching HMF_CI cookies. This is a common web scraping technique but introduces a dependency on the external API's cookie mechanism. Data and cookie caches are stored locally as JSON files. Input validation is implemented for date and count parameters to prevent malformed input. No hardcoded sensitive secrets or obvious malicious code patterns were found. File system operations are limited to cache management within the project directory.
Updated: 2025-12-05GitHub→
PreviousPage 147 of 713Next