The MCP Server Directory
Every Model Context Protocol server, scored on reliability, documentation, maintenance, and adoption. Independent, opinionated, refreshed daily from the official registry.
What this is
The MCP Server Directory is a live, independently maintained index of Model Context Protocol servers published to the official registry. Each server is scored 0–10 on four dimensions: reliability, documentation quality, maintenance velocity, and adoption. The directory is free, updated daily, and built by MegaOne AI.
Voicegen Port
4.1/10Generate speech, voices, and transcripts through the ElevenLabs API.
Nextjs Seo Audit
4.1/10Read-only SEO and indexing audit of a deployed site, aimed at Next.js
Pdfstudio
4.1/10Read, search, edit, and truly redact PDFs locally — 10 agent tools over stdio.
Podkop MCP
4.1/10MCP server for managing Podkop and Forkop on OpenWrt over SSH
Nix
4.1/10Servidor MCP stdio para wikis em Markdown
Actionbridge
4.1/10Zapier exposed actions as MCP tools via a generic action runner.
Aidocs
4.1/10OpenAI docs, guides, changelog and API reference as MCP tools.
Apphost Port
4.1/10Render services, deploys, env vars and logs as MCP tools.
Ashareport
4.1/10A-share stock data: history, realtime, news, statements and metrics.
Calendarport
4.1/10Google Calendar calendars, events and free-busy checks, as MCP tools.
Channelchat
4.1/10Slack channels, messages, threads, reactions and users via the Slack Web API.
Cloudinfra
4.1/10Terraform Registry docs, module search and AWS plan policy scans as MCP tools.
Cloudkb
4.1/10Bedrock knowledge bases, data sources and retrieval as MCP tools.
Codehost Port
4.1/10GitHub repositories, issues and pull requests via the REST API, as MCP tools.
Codehost Remote
4.1/10Forward GitHub tools to the hosted GitHub Copilot remote MCP endpoint.
Collabnotes
4.1/10Collaborative notes via HackMD: user, team, and personal notes with full CRUD.
Crmport
4.1/10HubSpot CRM objects, properties, owners, and associations: list, search, write.
Cryptoprice
4.1/10CoinGecko prices, coin data, search, charts, global and trending data, keyless.
Deployport
4.1/10Vercel projects, deployments, domains and env vars as MCP tools.
Diagramport
4.1/10Generate Mermaid diagrams and charts from Mermaid syntax, validated locally.
Docxredline
4.1/10Read, redline, and finalize .docx files with native Word tracked changes.
Errorwatch
4.1/10Organizations, projects, releases, and issues through the Sentry API.
Forumport
4.1/10Reddit subreddits, posts, comments and users as MCP tools.
Imageport
4.1/10Local image tools via ImageMagick: metadata, resize, convert, crop, rotate.
Issueport
4.1/10Linear teams, projects, issues and comments via GraphQL, as MCP tools.
Localmodels
4.1/10Talk to a local Ollama server: list, pull, inspect, generate, chat and embed.
Modelhub Port
4.1/10Search and inspect Hugging Face models, datasets, spaces, papers and docs.
Modelrun
4.1/10Replicate models, collections, versions, and predictions: browse and run.
Musicport
4.1/10Spotify search, playback and playlists as MCP tools.
Nasport
4.1/10Unraid server operations router: system, array, Docker, VMs and more.
Pageworkspace
4.1/10Notion pages, databases and blocks via the REST API, as MCP tools.
Productmetrics
4.1/10PostHog projects, insights, flags and experiments as MCP tools.
Recordcrm
4.1/10Attio CRM records, companies, deals, lists, and members via the Attio API.
Refdocs
4.1/10Search technical documentation and read web pages as markdown.
Serverlessrun
4.1/10AWS Lambda functions, versions, aliases, layers and settings as MCP tools.
Sheetport
4.1/10Google Sheets spreadsheets and cell values via the Sheets API, as MCP tools.
Sitehost Port
4.1/10Netlify sites, deploys and builds as MCP tools.
Slideport
4.1/10Google Slides presentations and slides via the Slides API, as MCP tools.
Stackconstruct
4.1/10AWS CDK guidance, docs, constructs and nag checks as MCP tools.
Supportchat
4.1/10Intercom conversations, contacts, tags and articles as MCP tools.
Taskboard Port
4.1/10Asana workspaces, projects and tasks via the REST API, as MCP tools.
Tasklist Port
4.1/10Todoist tasks, projects, labels and sections as MCP tools.
Videohub Port
4.1/10YouTube videos, channels, playlists, comments and transcripts as MCP tools.
Worktrack Port
4.1/10Jira issues and Confluence pages via the REST APIs, as MCP tools.
Muscle Memory
4.1/10MCP server that gives your agent a library of its own tested, reusable scripts.
MCP Server
4.1/10Meter usage and manage usage-based billing from AI tools via the Clocknext API.
Agentcomms
4.1/10Agent email, agent-to-agent messaging, and cloud file storage for AI agents.
Backportcheck
4.1/10Decide whether a fix needs a backport, and to which release branches.
Free · No spam
New servers, score changes, breakages
Every Monday: what shipped, what broke, what's worth trying. Five minutes. No filler.
Four dimensions, weighted
- Reliability (30%) — repository present, permissive license, not archived, official status.
- Documentation (20%) — description length, website presence, README quality.
- Maintenance (25%) — how recently the code was pushed; versioning discipline.
- Adoption (25%) — GitHub stars (log-scaled), remote availability, official endorsement.
What is MCP?
MCP stands for Model Context Protocol. It is an open standard introduced by Anthropic in late 2024 that lets AI assistants like Claude connect to external tools, data sources, and services through a common interface. An MCP server is a small program that exposes one capability (file access, a database, an API) to any MCP-compatible AI client.
Where does the data come from?
Every server in this directory is pulled from the official Model Context Protocol Registry at registry.modelcontextprotocol.io. We also enrich each entry with GitHub metadata (stars, last commit, license, primary language) where a repository is linked. The full re-scan runs daily.
How are servers scored?
Each server gets a 0–10 overall score, weighted from four sub-scores: reliability (30%), documentation (20%), maintenance (25%), and adoption (25%). The full methodology is visible on this page under "How scoring works".
Is this affiliated with Anthropic?
No. MegaOne AI is independent. Anthropic created the MCP standard and maintains the reference implementations, but this directory is built separately and rates all servers including Anthropic's own.
How often does the directory update?
The pipeline re-fetches the full registry and re-scores every active server once per day. Individual entries are refreshed whenever a new version is published to the official registry.
Is it free?
Yes. Browsing, filtering, and searching the directory is free. The weekly email briefing is free. There is no account required.