Remote MCP servers
43.730 servers, die du per URL über streamable HTTP verbindest. Nichts zu installieren oder auf deiner Maschine laufen zu lassen.
Filter · 2
Vertrauen
Sprache
Kategorie
259 Ergebnisse
ServerCapabilitiesStars · 30d
43.730 servers, die du per URL über streamable HTTP verbindest. Nichts zu installieren oder auf deiner Maschine laufen zu lassen.
Füge das GitHub Repo ein und wir listen ihn nach einem kurzen Review auf.
Server einreichenEin Remote MCP server läuft auf fremder Infrastruktur und dein Client verbindet sich per URL über streamable HTTP (ältere Server nutzen SSE). Die Anmeldung läuft meist über OAuth im Browser statt über einen API key in einer Config-Datei.
Lokale Server laufen als Prozess auf deiner Maschine und erreichen deine Dateien und dein privates Netzwerk. Remote-Server brauchen keine Runtime und aktualisieren sich selbst, aber deine Daten gehen zum Anbieter. Nimm remote für SaaS-Produkte, lokal für Dateien, Datenbanken in deinem Netzwerk und alles Offline.
Führe in Claude Code claude mcp add --transport http <name> <url> aus. In Cursor und Gemini CLI fügst du die URL unter mcpServers hinzu; in VS Code unter servers mit "type": "http". Jede Server-Seite enthält die exakte Config.
Lets AI agents log conversation exchanges and recall persistent memory context through MCP tools.
$claude mcp add --transport http synap https://synap-mcp.maximem.ai/mcp
Provides a cross-agent intelligence network that shares tool failure patterns and successful recovery actions to prevent redundant retries.
$claude mcp add failecho -- npx -y failecho-mcp
Allows AI agents to create test inboxes and automatically block until an OTP or magic link arrives for application testing.
$claude mcp add mailsocket-mcp -- uvx mailsocket-mcp
Gives AI coding agents Figma design context so they can generate code, extract variables and components, and write to the canvas.
$claude mcp add --transport http mcp-server-guide https://mcp.figma.com/mcp
Compiles written policies into hashed rulesets so an AI client can extract claim values and check every assertion deterministically.
$claude mcp add dsail -- uvx dsail
Lets AI agents design, slice, and print 3D objects on connected printers, with Git-style version control for designs.
$claude mcp add kiln -- uvx kiln3d
Lets an AI client read Opik traces, log scores, and manage prompt versions in an LLM observability workspace.
$claude mcp add opik-mcp -- uvx opik-mcp
Lets an AI client run Google and other search engine queries and extract or crawl web page content via SerpApi.
$claude mcp add --transport http serpapi-mcp https://mcp.serpapi.com/mcp
Exposes Appwrite's API for databases, users, functions, teams, and storage as tools for AI clients to manage backends.
$claude mcp add mcp -- uvx mcp-server-appwrite
Enables AI clients to perform data loading, cleaning, visualization, and machine learning workflows using pandas-powered tools.
$claude mcp add stats-compass-mcp -- uvx stats-compass-mcp
Lets an AI client run DCF, NAV, WACC, options, credit risk, and actuarial valuations with IVS and IFRS citations.
$claude mcp add fair-value -- uvx fair-value
Provides a voice interface for Claude Code, enabling speech recognition and spoken replies during agentic coding tasks.
$claude mcp add noisy-coding -- docker run -i --rm docker.io/noisy/noisy-coding:2.13.3
Provides structured plan, build, review, and release cycles for AI coding assistants to maintain project direction and decision history.
$claude mcp add papi -- npx -y @papi-ai/server
Indexes repository markdown to provide coding agents with a single canonical document per query, reducing token usage by approximately 60%.
$claude mcp add trovex -- uvx trovex
Lets AI assistants perform IP intelligence lookups including geolocation, network ownership, and privacy detection.
$claude mcp add mcp -- uvx ipinfo-mcp-server
Enables AI assistants to search electronic components, compare prices, check availability, and process BOMs from various EDA tools.
$claude mcp add parts-mcp -- uvx parts-mcp
Lets an AI client list, inspect, and grant or deny pending CTRLRun approval requests.
$claude mcp add ctrlrun -- uvx ctrlrun
Lets an AI client run allowlisted, audited commands and structured file edits on Linux, macOS, and Windows hosts.
$claude mcp add --transport http sentinelx-cloud-core https://mcp.sentinelx.app/mcp/mcp
Provides a formal verification engine to test AI claims against ground-truth axiom libraries to eliminate hallucinations and return binary verdicts.
$claude mcp add synproof-syntropy -- npx -y syntropy-mcp
Provides structured failure knowledge to prevent AI agents from repeating known code errors and country-specific real-world mistakes.
$claude mcp add deadends-dev -- uvx deadends-dev