Every Grade-A MCP server
7,604 of the 27,805 Model Context Protocol servers scanned by the Canopii Trust Index score 90–100/100 — a security grade of A, or 27% of the corpus.
These servers pass every security control that applies to them, with enough of their source verifiable for us to say so. No confirmed injection sinks, no committed credentials, no known-vulnerable dependencies left unpatched.
- Average score
- 92/100
- Share of scanned corpus
- 27%
- Verified against a live endpoint
- 447
The 100 highest-scoring
Ranked by score, then by how much of the server we could verify.
Mock, record/replay, verify and chaos-test any HTTP, REST, gRPC or LLM dependency over MCP.
An MCP (Model Context Protocol) server implementation for Microsoft Teams integration, providing capabilities to read messages, create messages, reply to messages, mention members.
Run and manage H Company's Computer-Use Agents from any MCP client.
Local-first memory for AI agents about the people in your life. MCP server + CLI on SQLite. Never phones home.
Manage Coder workspaces, templates, and cloud development environments
Rust release tooling — drop-in GoReleaser parity + Rust extras.
AI-driven penetration testing - 22 security tools behind safety-hardened MCP endpoints
MCP gateway for controlled SSH access with per-client auth, command policies, and audit logging.
Local verification of nomue Protocol Records; current release supports the Release 1 Welch bundle.
Secret scanning for AI agents: rules and entropy find candidates, a model decides which are real.
Discover open-source AI agents and get source-owned installation instructions through MCP.
Open-source local MCP connector for intervals.icu (Go binary)
Local Go repository call graphs and structural analysis for coding agents over stdio.
Query WHOIS/RDAP information for domains, IP addresses, CIDR prefixes and ASNs. Self-hostable.
Automate tasks, processes, and approvals with AI.
A Model Context Protocol (MCP) application for automated GitHub PR analysis and issue management.…
Stateless governance proxy for MCP (2026-07-28): policy, dry-run, human confirmation, audit, tracing
Read-only tools to inspect your agent setup: list MCP servers, flag unpinned ones, test a policy.
Read and write a self-hosted TriliumNext knowledge base over its ETAPI.
A lightweight Model Context Protocol (MCP) server that enables AI assistants like Claude to retrieve and interpret real-time weather data. Discuss on Hacker News:
Knowledge graph for token-efficient code reviews -- semantic search and call-graph resolution across your codebase.
Search, resume and convert AI coding sessions (Claude Code, Codex, Gemini, OpenCode, Pi) over MCP.
Search, read, and edit any MediaWiki wiki. Works with Wikipedia, Fandom, and more.
🚀 Self-hosted TikTok & Douyin scraper and no-watermark video downloader — async REST API, MCP server, CLI and web console for posts, profiles, comments and playlists. Self-healing identity pool, PostgreSQL archive, one docker compose up. 抖音、TikTok 数据采集与无水印视频下载 API,自托管,支持 MCP 调用与 Docker 一键部署。
Discover and quote quality-gated private AI inference without creating or buying a locker.
Your Windows syscall hooking factory - feat Canterlot's Gate - All accessible over MCP
Turn Claude, Cursor, and Codex into web performance experts. A 16-tool MCP server for PageSpeed Insights, Lighthouse, CrUX, Core Web Vitals, audits, diagnostics, and optimization.
Python MCP Server for Kestra — you can use it as a tool in Kestra's AI Agents
An MCP server retrieving transcripts of YouTube videos
ReActMCP is a reactive MCP client that empowers AI assistants to instantly respond with real-time, Markdown-formatted web search insights powered by the Exa API.
MCP server providing systematic thinking, mental models, and debugging approaches for enhanced LLM problem-solving
Read-only Kubernetes diagnostics: pods, logs, events, workloads, services, PVCs and nodes.
Real-time social media and prediction market integrity layer for traders, funds, and quant desks.
Connect your AI assistants to Keboola and expose your data, transformations, SQL queries, ...
Declarative agent orchestration for engineering teams
Read-only merchant data from 8 Chinese e-commerce platforms: orders, products, after-sales, ads
MCP server for Kylas CRM: create, search, and manage leads. Requires KYLAS_API_KEY.
A Model Context Protocol (MCP) server and app for creating Datawrapper charts using AI assistants
Local-first MCP server giving AI agents contextual knowledge about the people in your life
Real-time Google Flights: one-way and round-trip fares over date ranges, with price verdicts.
Live Booking.com hotel prices, plus per-country pricing for rate-parity monitoring.
Open source version of Claude Managed Agents. Fastest way to build and deploy reliable AI agents, MCP tools and agent-to-agent.
Reference verification MCP server for CSL bibliographies, evidence, and audited writes.
Bounded, source-write-free Rust correctness tools grounded in Cargo and rustc output.
Bounded, source-write-free Rust correctness tools grounded in Cargo and rustc output.
Security sidecar for MCP servers: 11 read-only control-plane tools for prompt-injection checks.
ESPN draft and lineup tools for Codex and other MCP clients, with browser sign-in, approval modes, per-team limits, and server monitoring.
Read-only MCP server for licensed Healthpoint HL7 FHIR API access.
CLI for exploring AWS EC2 Spot inventory. Inspect AWS Spot instance types, saving, price, and interruption frequency.
Remote MCP server for full read/write access to a Zotero library
Vedic and Western astrology for AI agents: charts, dasha, matchmaking, panchanga, numerology, tarot.
Unofficial MCP server for Leantime: 42 agent-first tools, backup/restore, dry-run validation
Notepatra editor MCP server — 35 tools; every write human-approved in the editor; local-only.
Manage Pi-hole v6: DNS blocking, domains, clients, query analysis, DHCP, and multi-instance sync.
Task-shaped Zabbix access for AI agents: problems, host triage, metrics, maintenance windows.
The IDE for agents: read by symbol, edit against a revision, validate with your own build, revert.
comprehensive Discord selfbot MCP server using discord.py-self
MCP server for guiding Coding Agents via end-to-end requirements to implementation plan pipeline
Autonomous log surgery & secret redaction MCP server for AI coding agents.
Create, edit, validate, preview, and render Shotcut projects through AI clients.
Search, move, mark, delete and draft your real mail over a self-hosted index.
Production-grade AutoCAD MCP server for AI agents — 122 tools, dual COM (live AutoCAD) + headless ezdxf engines, ISO GD&T and dimension-tolerance validation for CAD automation.
Read-only Kubernetes MCP server: inspect resources, logs, events, and metrics. Secrets are masked.
Generate more accurate Terraform and automate workflows for HCP Terraform and Terraform Enterprise
MCP server exposing rlg log streams to LLM agents via JSON-RPC 2.0 over stdio.
Task-shaped Zabbix access for AI agents: problems, host triage, metrics, maintenance windows.
🌍 Terraform Model Context Protocol (MCP) Tool - An experimental CLI tool that enables AI assistants to manage and operate Terraform environments. Supports reading Terraform configurations, analyzing plans, applying configurations, and managing state with Claude Desktop integration. ⚡️
Semantic code search and analysis from CodeAlive for AI assistants and agents.
Plant genomics MCP — 50 tools across 23 backends with cross-source synthesis.
Agent memory engine — 86 MCP tools, self-hosted, single Rust binary
Access Nordic business registries (Norway, Denmark, Finland, Sweden). 23 tools.
MCP server for C# code analysis and automated fixing using Roslyn analyzers and code fix providers.
The IDE for agents: read by symbol, edit against a revision, validate with your own build, revert.
An advanced sequential thinking process using a Multi-Agent System (MAS) built with the Agno framework and served via MCP.
Shared session bus MCP+CLI for Claude/Codex/Cursor/Grok: list, read, handoff, send, spawn.
Read-only access to a Crossplane control plane: resources, claims, packages, compositions.
A lightweight, Git-backed permanent memory for AI agents.
mcp-gitee is a Model Context Protocol (MCP) server implementation for Gitee. It provides a set of tools that interact with Gitee's API, allowing AI assistants to manage repository, issues, pull requests, etc.
Full Agent Swarm API and MCP server with local SQLite storage for multi-agent orchestration.
Local QMD search and bounded index maintenance through a Streamable HTTP MCP server.
MCP retrieval server for indexed Amazon Ads, Seller Central, and Vendor Central knowledge.
Control local KeyShot scenes, cameras, environments, and renders through headless scripting.
Claude 나 Codex 에서 벨로그(velog.io) 글을 읽고 쓰고 백업한다. 다이어그램과 시퀀스도 그린다. 공개 발행은 사람이 켤 때만 열린다.
MCP Server and CLI for Apache Spark History Server. Debug Spark applications from AI agents, scripts, or the terminal.
Buddhist canon tools: search, passages, cross-canon parallels, dictionaries — all URN-cited.
Cited, OSV-gated, injection-hardened web briefs for coding agents. Verified, never raw web.
Review coding-agent plans on demand with peer models, attributed findings, and a structured verdict.
An AI agent's safe read-only scout to orient in an unknown repository before touching it.
Import CSV/XLSX accounting data into QuickBooks Online or Xero via the Synder Importer API.
Analyze HOI4 events, tech trees, AI weights, and MTTH; build focus trees, GUIs, and maps.
MCP runtime copilot for health checks, regressions, explainability, and operational memory.
Local SwiftUI changed/staged checks with compact agent output and multi-module ownership.
MCP server that creates Amazon listing images and product video from live catalog data.
Anonymous public tools for FlightSweeper. See the published agent boundary before use.
Search your developer network and trace connection paths via GitHub & LinkedIn.
Route MCP capabilities through deterministic identity, policy, budget, and deadline constraints.
Multi-protocol API mock server: HTTP, GraphQL, gRPC, WebSocket, MQTT, SSE, SOAP.
Operate secure JoinLayer data pipelines through delegated OAuth.
Coding guidelines and best-practice guides for 130+ languages, frameworks, and tools.
Read and edit Overleaf papers from any AI client. Edits can arrive as tracked-change suggestions.
What a A actually means
Scores are deterministic and evidence-based: every point traces to a named control evaluated against the server's source, and a confirmed flaw caps the score outright regardless of what else passes. Grades are not endorsements and not a substitute for your own review — read the full methodology or see how the whole ecosystem scores.
Maintain one of these? Every server page carries an embeddable badge that updates on each rescan.