Updated weekly · Free & open
Which MCP servers are actually maintained?
The official registry lists every Model Context Protocol server but tells you nothing about whether any of them is still alive. This index scores each one on maintenance activity, flags what deserves a second look before you install it, and shows exactly how every number was reached.
45,870
servers indexed
87
with verified activity
20.0
median score
15,176
carrying risk flags
last refreshed
The weekly report
What moved, what dropped, and what appeared — a short read every Monday. No spam, unsubscribe in one click.
Leaderboard
Activity & maintenance, not security — how this is scored| # | Server | Flags | Stars | Change | Score |
|---|---|---|---|---|---|
| 11151 | MCP Dados Brasil Servidor MCP com dados oficiais brasileiros: IBGE, Banco Central, INMET, Câmara e Senado. | clean | — | — | 24.0 |
| 11152 | sota-anchor Finds what shipped after your coding agent's training, before it rebuilds it from scratch. | clean | — | — | 24.0 |
| 11153 | pyscn Python code analysis for AI agents: complexity, dead code, clones, coupling, and a health score. | clean | — | — | 24.0 |
| 11154 | Obsidian Vault MCP Local Zotero, MinerU, Obsidian literature workflow with transactional Analysis notes. | clean | — | — | 24.0 |
| 11155 | OmniScholar Literature, Zotero, PDF, citation, materials, and scientific image tools for research agents. | clean | — | — | 24.0 |
| 11156 | io.github.lugdwei/AgentMesh MCP delegation fallback for AI agents to discover capabilities, knowledge, tools, and collaborators. | clean | — | — | 24.0 |
| 11157 | PerspectiveGraph Read-only attack-path tools: reachable routes to sensitive assets, and what a fix would cut. | clean | — | — | 24.0 |
| 11158 | io.github.luisgf/infrabroker SSH & Kubernetes access broker for AI agents; the model never touches a credential. | clean | — | — | 24.0 |
| 11159 | closure_drift Before an agent tags a release: does that version label already name different code in the git tags? | clean | — | — | 24.0 |
| 11160 | io.github.lukabudik/alza-mcp Unofficial MCP server for Alza.cz: catalog search plus token-guarded account and checkout. | medium | — | — | 24.0 |
| 11161 | io.github.lukebmandel-debug/alto-connector Build an interactive timeline from your own materials — never from invented content. | clean | — | — | 24.0 |
| 11162 | CalDAV MCP Server A TypeScript MCP server for iCloud Calendar with native multi-VALARM support. | clean | — | — | 24.0 |
| 11163 | open-compute (computer-use) Model-agnostic computer-use MCP: screenshot, safety-gated actions, Windows UIA targeting. | clean | — | — | 24.0 |
| 11164 | io.github.lumayapartners/memini Local-first project memory and mistake-prevention guardrails for AI coding agents. | clean | — | — | 24.0 |
| 11165 | PostSider Social media calendar for AI agents: read channels, create drafts, request approval. | clean | — | — | 24.0 |
| 11166 | io.github.luno-cms/mcp LUNO — AI-era Backend Platform (not Luno crypto). BUILD/OPERATE, REGISTER/GOVERN, CONSUME. | clean | — | — | 24.0 |
| 11167 | echorune radar Live weather anywhere, with the radar drawn as text characters, not an image. | clean | — | — | 24.0 |
| 11168 | EasyAccessPDF PDF accessibility checks (veraPDF PDF/UA-1), auto-fix and Markdown conversion. EU-hosted. | clean | — | — | 24.0 |
| 11169 | io.github.lupingQAQ/ntobjmanager-mcp Stateful Windows RPC attack-surface research for AI agents: 22 tools on NtObjectManager. | clean | — | — | 24.0 |
| 11170 | io.github.lusiem/code-atlas Code intelligence MCP: symbols, call graphs, change impact + affected tests, routes, search. | clean | — | — | 24.0 |
| 11171 | Keel Control plane MCP that dedups scanner output, rate-limits hosts, and runs non-destructive proofs. | clean | — | — | 24.0 |
| 11172 | io.github.lvlrSajjad/sidecrew Code changes and unit tests by local MLX models, behind a gate a machine can run. | clean | — | — | 24.0 |
| 11173 | simframe Always-warm iOS Simulator and Android emulator frames: agents read the screen in ~20ms | clean | — | — | 24.0 |
| 11174 | Xcatcher — Recent X Posts Fetch recent public X/Twitter posts by named handle for monitoring, comparison, OSINT, and research. | clean | — | — | 24.0 |
| 11175 | OBSBOT Tiny 2 Camera Control Control an OBSBOT Tiny 2 camera: pan/tilt/zoom, AI tracking, focus, exposure, snapshots, recording | clean | — | — | 24.0 |
| 11176 | io.github.lxxmng/container-loading Plan optimal container & truck loads: 3D layouts, utilization, centre of gravity, crush checks. | clean | — | — | 24.0 |
| 11177 | io.github.lxxmng/ocean-schedules Compare ocean carrier schedules, transit times and reliability: 60+ carriers, 72,000+ sailings. | clean | — | — | 24.0 |
| 11178 | io.github.lyf-wxy/spomory Explainable graph-retrieval memory engine with an RL-trained management policy. | clean | — | — | 24.0 |
| 11179 | U.S. Government Open Data Open data from 40+ U.S. government APIs: economy, spending, Congress, health, energy, disasters | clean | — | — | 24.0 |
| 11180 | Zektor Managed Postgres and Valkey hosted in Germany. Agents can create a trial database with no signup. | clean | — | — | 24.0 |
| 11181 | Calibration Explorer Check classifier confidence against accuracy, fit temperature scaling and save local reports. | clean | — | — | 24.0 |
| 11182 | M00N Report Test management over MCP: author cases, run manual executions, cut releases, read health | clean | — | — | 24.0 |
| 11183 | io.github.m0rvayne/mcp-osascript macOS automation: 17 typed tools for windows, menus, keyboard, clipboard, tabs, screenshots | clean | — | — | 24.0 |
| 11184 | io.github.m9labs-railscope/verantis-mcp Verified directory for machine payments (x402 & MPP): check services and wallets before agents pay. | clean | — | — | 24.0 |
| 11185 | Outlook Mailbox Tidier Tidy a large Outlook / Hotmail mailbox. Folder subtree moves, bulk ops with dry run, never sends. | clean | — | — | 24.0 |
| 11186 | Contexter Share one project context across ChatGPT, Claude, Telegram and any MCP client. | clean | — | — | 24.0 |
| 11187 | Jaxonomy — hybrid dynamical systems on JAX Build, simulate and calibrate JAX block-diagram models of hybrid dynamical systems. | clean | — | — | 24.0 |
| 11188 | io.github.machinegrade/validate Deterministic validation for AI-generated artifacts: JSON Schema, OpenAPI response, SQL syntax. | clean | — | — | 24.0 |
| 11189 | io.github.mack-peng/cssgraph CSS intelligence for AI coding agents — surgical style context, fewer tool calls. 100% local. | clean | — | — | 24.0 |
| 11190 | io.github.mack-peng/cssprobe-cli Runtime CSS probe — inspect layout, overflow, scroll chains in a real browser. For AI agents. | clean | — | — | 24.0 |
| 11191 | Bancadia MCP Query Bancadia's registry of US business checking account products with structured filters. | clean | — | — | 24.0 |
| 11192 | io.github.macuse-app/macuse MCP server for Mac — control Calendar, Mail, Notes, Reminders, and any Mac app, fully on-device. | clean | — | — | 24.0 |
| 11193 | comdirect Read-only access to comdirect accounts, transactions, depots, orders and postbox documents. | clean | — | — | 24.0 |
| 11194 | Nexus MCP Search Notion and Supabase through one safety-focused MCP server. | clean | — | — | 24.0 |
| 11195 | io.github.madjik-io/mcp-server Quantum-finance market metrics for AI agents, computed by the Black Belt Labs engine. | clean | — | — | 24.0 |
| 11196 | Onchain Data MCP Blockchain data for EVM and Solana: balances, tokens, prices, DeFi and RPC via free-tier vendors | clean | — | — | 24.0 |
| 11197 | proactive-mcp Detects reply deadlines, calendar conflicts, and personal occasions for MCP agents. | clean | — | — | 24.0 |
| 11198 | Zamery Browser Use your existing Firefox and logged-in tabs, limited to tabs or groups you explicitly share. | clean | — | — | 24.0 |
| 11199 | io.github.mafedelahoz/cuddlynest-mcp Search CuddlyNest hotels and read rooms, prices, availability and cancellation policies | medium | — | — | 24.0 |
| 11200 | Casper Tools 19 zero-setup utilities: JSON, regex, cron, hash, base64, CSV, JWT, UUID, YAML, semver, time. | clean | — | — | 24.0 |
A score measures maintenance activity, not safety. A high score does not mean a server is secure, and a low one does not mean it is dangerous — see the methodology for what each signal does and does not tell you.