Annuaire des serveurs MCP
56,350 serveurs MCP qui relient les assistants IA à des outils et des données, issus du registre MCP officiel et des listes communautaires, mis à jour toutes les heures.
- 16 catégories
- 36,250 serveurs distants
- Mis à jour toutes les heures
com.mylinedchart/chart-contextcom.mylinedchart/chart-context
Read-only MCP server: AI agents read your live MyLinedChart desktop chart context. Local, IBKR.
io.github.Bishop81/accessibility-scanner-mcpio.github.Bishop81/accessibility-scanner-mcp
Scan any URL for WCAG accessibility issues (axe-core) and get per-element fixes for an AI agent.
io.github.abhiyoheswaran1/baseframelabs-catalogueio.github.abhiyoheswaran1/baseframelabs-catalogue
Agent-native catalogue of Baseframe Labs dev tools and MCP servers.
io.github.sachinuppal/modelcostsaverio.github.sachinuppal/modelcostsaver
Offline MCP: predict LLM call cost and pick the cheapest capable model. No keys, no network.
io.github.MS-Teja/glyphicio.github.MS-Teja/glyphic
Generate 18 diagram types from structured JSON natively. Built for LLMs and agents.
xyz.fluxiq/agentxyz.fluxiq/agent
Non-custodial Ethereum swaps and managed liquidity. You sign every tx; FluxIQ holds no funds.
io.github.Michael-WhiteCapData/ollama-handoffio.github.Michael-WhiteCapData/ollama-handoff
Offload cheap work from your AI agent to a local Ollama model, at zero cloud cost.
io.github.FoundryNet/foundrynet-inferenceio.github.FoundryNet/foundrynet-inference
x402 LLM proxy + data-enriched analysis (17 sources) + TimesFM predictive IoT intelligence.
com.claritybriefs/mcpcom.claritybriefs/mcp
Free cross-lingual news briefings for AI agents across 89 languages. Read-only, hosted.
io.github.medelin/bashdog-mcpio.github.medelin/bashdog-mcp
Submit & browse funny AI quotes on bash.dog — bash.org for the AI age. Agents self-register.
com.ledgenter/mcpcom.ledgenter/mcp
The shared work-management office for AI agents — projects, tasks, decisions, knowledge, handoffs.
io.github.waqarulwahab/llm-cost-estimatorio.github.waqarulwahab/llm-cost-estimator
Token counting & multi-model LLM cost estimates: GPT-4o, Claude, Gemini, 25+. No API key.
io.github.birthday-tools/edgarmcpio.github.birthday-tools/edgarmcp
Clean, normalized SEC EDGAR / FRED / ETF / index data for AI agents.
io.github.0xChron/polymarket-mcpio.github.0xChron/polymarket-mcp
Read-only Polymarket prediction market data for AI agents.
io.github.derekchoyai/honeio.github.derekchoyai/hone
Score your judgment over AI-assisted work. Hone scores you, not the model.
io.github.Goldentrii/agent-recallio.github.Goldentrii/agent-recall
Correction-first agent memory. Precision KPI tracks if agents heed warnings. 5 layers, local-only.
ITHZ MCPdev.ithz/ithz-mcp
Local deterministic project memory for AI agents with context packs, gates and checkpoints.
io.github.douglasgan/asktianio.github.douglasgan/asktian
Chinese metaphysics (bazi, qimen, 5-element) as decision-support tools for AI agents.
io.github.mbtest870-ctrl/agenzaio.github.mbtest870-ctrl/agenza
AI agent reputation on ICP.
ai.spideriq/gateai.spideriq/gate
SpiderIQ Gate: SpiderGate LLM gateway (completions, models, usage, traces)
ai.spideriq/mailai.spideriq/mail
SpiderIQ Mail: SpiderMail mailboxes, messages, threads, send, compose for AI agents
io.github.datasynx/cartographyio.github.datasynx/cartography
MCP-first read-only discovery of your infra & SaaS landscape as MCP resources, tools and prompts.
io.github.qinisolabs/companieswiseio.github.qinisolabs/companieswise
Verified UK company lookup & validation for AI agents — Companies House data, snapshot or live.
com.llm-bus/llm-buscom.llm-bus/llm-bus
Coordinate multiple AI agents over MCP: atomic claims, leases, shared ledger, handoffs, tasks.
online.claudenews/claude-newsonline.claudenews/claude-news
Real-time Claude & Anthropic news from your own Claude — free, no API key.
RZ AI Labscom.rzailabs/rzailabs-mcp
Read-only MCP server for RZ AI Labs — query its services, workshops, and contact info.
io.github.qinisolabs/sanctionwiseio.github.qinisolabs/sanctionwise
Verified UK sanctions screening for AI agents — official FCDO UK Sanctions List, not guesses.
io.github.dhnpmp-tech/dcp-mcpio.github.dhnpmp-tech/dcp-mcp
Saudi sovereign AI compute: OpenAI-compatible inference, rent GPU pods, manage storage.
Wever Labs Agentic Railscom.weverlabs/agentic-rails
Agentic rails for complex workflows with receipts, fees, and MCP tool access.
io.github.us-all/mlflowio.github.us-all/mlflow
MLflow MCP — experiments, runs, registered models, versions, traces, assessments (MLflow 3)
io.github.qinisolabs/floodwiseio.github.qinisolabs/floodwise
England flood-risk by postcode for AI agents — verified Environment Agency data, not guesses.
io.github.xfloukiex-lab/toki-tokiio.github.xfloukiex-lab/toki-toki
A local causality ledger: records decisions and the reasons, for an agent to recall.
io.github.kotinder/roomcommio.github.kotinder/roomcomm
Ephemeral REST chatrooms for AI agents to coordinate. Share a room URL — agents talk live.
io.github.sara-star-quant/presence-mcpio.github.sara-star-quant/presence-mcp
Read-only MCP access to presence: living project model and outcome telemetry.
io.github.silentdany/brieformio.github.silentdany/brieform
AI-native form builder: create, publish & read responses from Claude, ChatGPT & MCP.
io.github.whazeted/paraglide-messages-mcpio.github.whazeted/paraglide-messages-mcp
Validated, sub-agent per-locale translation of Paraglide JS / inlang message-format projects.
io.github.rccola990-cloud/mcp-token-optimizerio.github.rccola990-cloud/mcp-token-optimizer
Cut LLM token costs: count tokens, estimate cost, slim prompts, and pick the cheapest capable model.
io.github.helinakdogan/agent-magnetio.github.helinakdogan/agent-magnet
Behavioral memory layer for AI tools. Learns preferences and context across Claude, Cursor, Codex.
Ultra Promptco.ultraprompt/ultra-prompt
Your prompt library inside your AI: 1,000+ pro templates, frameworks, vocab & pipelines.
io.github.Jolymmiles/fusion-mcpio.github.Jolymmiles/fusion-mcp
Local Fusion MCP: panel->judge->synthesis on the Claude CLI subscription, no OpenRouter.
com.zeroheight/zeroheightcom.zeroheight/zeroheight
Connect AI assistants to zeroheight styleguides using the Model Context Protocol
io.github.qinisolabs/bizdaysio.github.qinisolabs/bizdays
Verified business-day math for AI agents: holidays + per-country weekends, not guesses.
io.github.Daichi-Kudo/llm-advisorio.github.Daichi-Kudo/llm-advisor
Real-time LLM/VLM benchmarks, pricing, and recommendations. 300+ models, 5 sources.
io.github.AkynZH/pash-mcp-serverio.github.AkynZH/pash-mcp-server
Standalone MCP server with PASH compression. Saves ~87% LLM tokens via dynamic tool registration.
io.converly/converlyio.converly/converly
Set up and manage conversion tracking by chatting with your AI assistant.
io.github.MrWizardlyLoaf/rugcheck-aiio.github.MrWizardlyLoaf/rugcheck-ai
Solana token safety for AI agents — rug-pull, honeypot & Token-2022 trap detection before you buy.
io.github.Evin009/smartrent-mcpio.github.Evin009/smartrent-mcp
Control SmartRent smart home devices (locks, thermostat, switches, sensors) via Claude
io.github.geraservicesuk/mcp-gera-witnessio.github.geraservicesuk/mcp-gera-witness
GeraWitness MCP (placeholder) — human-in-the-loop safety layer for AI agents. Launching 2026.
Issu du registre MCP officiel. Être listé n'est pas une recommandation : vérifiez le code et les permissions d'un serveur avant de lui donner accès à vos données. Dernière synchronisation 2 nov. 2026.
Questions fréquentes
Qu’est-ce que MCP ?
Le Model Context Protocol est un standard ouvert qui permet à un assistant IA (Claude, ChatGPT, Cursor, un agent) d'appeler des outils et de lire des données externes via un « serveur ». Un serveur distant est une URL à ajouter à votre client ; un serveur local tourne sur votre ordinateur à partir d'un paquet (npm, PyPI, Docker).
Comment ajouter un serveur MCP à Claude ou Cursor ?
Ouvrez la page du serveur ici, copiez son bloc de configuration dans le fichier de réglages MCP de votre client, puis redémarrez-le. Un serveur distant n'a besoin que de l'URL ; un serveur local demande Node (npx), Python (uvx) ou Docker.
Ces serveurs sont-ils sûrs ?
N'importe qui peut publier dans le registre. Un serveur peut lire ce à quoi vous lui donnez accès et agir en votre nom : préférez ceux dont le code est public et l'éditeur reconnu, et accordez le minimum de droits.