[{"_1":2,"_305":-5,"_306":-5},"loaderData",{"_3":4,"_22":23,"_24":25,"_250":251},"root",{"_5":6,"_7":8},"toasts",[],"visitor",{"_9":10,"_11":12},"viewportWidth",1280,"visitorSession",{"_13":14,"_16":17,"_18":-5,"_19":20,"_21":-5},"attributes",[15],"BOT","id",6120031,"membership","referenceId","0196b156-ebbf-706b-a975-cde13b11758f","userAccount","routes/_public/_layout",{},"routes/_public/mcp/servers/~namespace/~slug/_pages/_layout",{"_26":27,"_82":-5,"_83":40,"_29":84,"_167":168,"_249":128},"availableLocales",[28,40,47,54,61,68,75],{"_29":30,"_33":37,"_38":39},"mcpServer",{"_31":32,"_33":34,"_35":36},"description","Ein auf Cloudflare Workers basierender Model Context Protocol-Server, der KI-Assistenten wie Claude den Zugriff auf externe Tools über die OAuth-Authentifizierung ermöglicht.","name","Remote-MCP-Server","readme","# Remote-MCP-Server auf Cloudflare\n\nLassen Sie uns einen Remote-MCP-Server auf Cloudflare Workers zum Laufen bringen, komplett mit OAuth-Anmeldung!\n\n## Lokal entwickeln\n\n```bash\n# clone the repository\ngit clone git@github.com:cloudflare/ai.git\n\n# install dependencies\ncd ai\nnpm install\n\n# run locally\nnpx nx dev remote-mcp-server\n```\n\nSie sollten [`http://localhost:8787/`](http://localhost:8787/) in Ihrem Browser öffnen können.\n\n## Verbinden Sie den MCP-Inspektor mit Ihrem Server\n\nUm Ihre neue MCP-API zu erkunden, können Sie den [MCP Inspector](https://modelcontextprotocol.io/docs/tools/inspector) verwenden.\n\n* Starten Sie es mit `npx @modelcontextprotocol/inspector`\n* Ändern Sie [im Inspektor](http://localhost:5173) den Transporttyp in `SSE` und geben Sie `http://localhost:8787/sse` als URL des MCP-Servers ein, mit dem eine Verbindung hergestellt werden soll. Klicken Sie anschließend auf „Verbinden“.\n* Sie gelangen zu einem (simulierten) Anmeldebildschirm mit Benutzername und Passwort. Geben Sie zur Anmeldung eine beliebige E-Mail-Adresse und Ihr Passwort ein.\n* Sie sollten zurück zum MCP-Inspektor umgeleitet werden und können nun alle definierten Tools auflisten und aufrufen!\n\n## Verbinden Sie Claude Desktop mit Ihrem lokalen MCP-Server\n\nDer MCP-Inspektor ist großartig, aber wir möchten ihn unbedingt mit Claude verbinden! Folgen Sie [der Schnellstartanleitung von Anthropic](https://modelcontextprotocol.io/quickstart/user) und gehen Sie in Claude Desktop zu Einstellungen > Entwickler > Konfiguration bearbeiten, um Ihre Konfigurationsdatei zu finden.\n\nÖffnen Sie die Datei in Ihrem Texteditor und ersetzen Sie sie durch diese Konfiguration:\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"http://localhost:8787/sse\"\n ]\n }\n }\n}\n```\n\nDadurch wird ein lokaler Proxy ausgeführt und Claude kann über HTTP mit Ihrem MCP-Server kommunizieren\n\nWenn Sie Claude öffnen, sollte sich ein Browserfenster öffnen, in dem Sie sich anmelden können. Unten rechts sollten Sie die verfügbaren Tools sehen. Bei entsprechender Eingabeaufforderung sollte Claude den Aufruf des Tools anfordern.\n\n## Bereitstellung auf Cloudflare\n\n1. `npx wrangler kv namespace create OAUTH_KV`\n2. Befolgen Sie die Anleitung, um die KV-Namespace-ID zu `wrangler.jsonc` hinzuzufügen\n3. `npm run deploy`\n\n## Rufen Sie Ihren neu bereitgestellten Remote-MCP-Server von einem Remote-MCP-Client aus auf\n\nFühren Sie den MCP-Inspektor aus, genau wie Sie es oben unter „Lokal entwickeln“ getan haben:\n\n`npx @modelcontextprotocol/inspector@latest`\n\nGeben Sie dann die `workers.dev` -URL (z. B. `worker-name.account-name.workers.dev/sse` ) Ihres Workers im Inspector als URL des MCP-Servers ein, mit dem eine Verbindung hergestellt werden soll, und klicken Sie auf „Verbinden“.\n\nSie haben jetzt von einem Remote-MCP-Client aus eine Verbindung zu Ihrem MCP-Server hergestellt.\n\n## Verbinden Sie Claude Desktop mit Ihrem Remote-MCP-Server\n\nAktualisieren Sie die Claude-Konfigurationsdatei, sodass sie auf Ihre `workers.dev` -URL verweist (z. B. `worker-name.account-name.workers.dev/sse` ), und starten Sie Claude neu.\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"https://worker-name.account-name.workers.dev/sse\"\n ]\n }\n }\n}\n```\n\n## Debuggen\n\nSollte etwas schiefgehen, kann es hilfreich sein, Claude neu zu starten oder zu versuchen, mit dem folgenden Befehl über die Befehlszeile eine direkte Verbindung zu Ihrem MCP-Server herzustellen.\n\n```bash\nnpx mcp-remote http://localhost:8787/sse\n```\n\nIn einigen seltenen Fällen kann es hilfreich sein, die zu `~/.mcp-auth` hinzugefügten Dateien zu löschen\n\n```bash\nrm -rf ~/.mcp-auth\n```","German","tag","de-DE",{"_29":41,"_33":45,"_38":46},{"_31":42,"_33":43,"_35":44},"A Cloudflare Workers-based Model Context Protocol server that enables AI assistants like Claude to access external tools via OAuth authentication.","Remote MCP Server","# Remote MCP Server on Cloudflare\n\nLet's get a remote MCP server up-and-running on Cloudflare Workers complete with OAuth login!\n\n## Develop locally\n\n```bash\n# clone the repository\ngit clone git@github.com:cloudflare/ai.git\n\n# install dependencies\ncd ai\nnpm install\n\n# run locally\nnpx nx dev remote-mcp-server\n```\n\nYou should be able to open [`http://localhost:8787/`](http://localhost:8787/) in your browser\n\n## Connect the MCP inspector to your server\n\nTo explore your new MCP api, you can use the [MCP Inspector](https://modelcontextprotocol.io/docs/tools/inspector).\n\n- Start it with `npx @modelcontextprotocol/inspector`\n- [Within the inspector](http://localhost:5173), switch the Transport Type to `SSE` and enter `http://localhost:8787/sse` as the URL of the MCP server to connect to, and click \"Connect\"\n- You will navigate to a (mock) user/password login screen. Input any email and pass to login.\n- You should be redirected back to the MCP Inspector and you can now list and call any defined tools!\n\n

\n $\"MCP$ \n

\n\n

\n $\"MCP$ \n

\n\n## Connect Claude Desktop to your local MCP server\n\nThe MCP inspector is great, but we really want to connect this to Claude! Follow [Anthropic's Quickstart](https://modelcontextprotocol.io/quickstart/user) and within Claude Desktop go to Settings > Developer > Edit Config to find your configuration file.\n\nOpen the file in your text editor and replace it with this configuration:\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"http://localhost:8787/sse\"\n ]\n }\n }\n}\n```\n\nThis will run a local proxy and let Claude talk to your MCP server over HTTP\n\nWhen you open Claude a browser window should open and allow you to login. You should see the tools available in the bottom right. Given the right prompt Claude should ask to call the tool.\n\n

\n $\"Clicking$ \n

\n\n

\n $\"Claude$ \n

\n\n## Deploy to Cloudflare\n\n1. `npx wrangler kv namespace create OAUTH_KV`\n2. Follow the guidance to add the kv namespace ID to `wrangler.jsonc`\n3. `npm run deploy`\n\n## Call your newly deployed remote MCP server from a remote MCP client\n\nJust like you did above in \"Develop locally\", run the MCP inspector:\n\n`npx @modelcontextprotocol/inspector@latest`\n\nThen enter the `workers.dev` URL (ex: `worker-name.account-name.workers.dev/sse`) of your Worker in the inspector as the URL of the MCP server to connect to, and click \"Connect\".\n\nYou've now connected to your MCP server from a remote MCP client.\n\n## Connect Claude Desktop to your remote MCP server\n\nUpdate the Claude configuration file to point to your `workers.dev` URL (ex: `worker-name.account-name.workers.dev/sse`) and restart Claude \n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"https://worker-name.account-name.workers.dev/sse\"\n ]\n }\n }\n}\n```\n\n## Debugging\n\nShould anything go wrong it can be helpful to restart Claude, or to try connecting directly to your\nMCP server on the command line with the following command.\n\n```bash\nnpx mcp-remote http://localhost:8787/sse\n```\n\nIn some rare cases it may help to clear the files added to `~/.mcp-auth`\n\n```bash\nrm -rf ~/.mcp-auth\n```\n","English","en-US",{"_29":48,"_33":52,"_38":53},{"_31":49,"_33":50,"_35":51},"Un servidor de protocolo de contexto de modelo basado en Cloudflare Workers que permite a los asistentes de IA como Claude acceder a herramientas externas a través de la autenticación OAuth.","Servidor MCP remoto","# Servidor MCP remoto en Cloudflare\n\n¡Pongamos en funcionamiento un servidor MCP remoto en Cloudflare Workers con inicio de sesión OAuth incluido!\n\n## Desarrollar localmente\n\n```bash\n# clone the repository\ngit clone git@github.com:cloudflare/ai.git\n\n# install dependencies\ncd ai\nnpm install\n\n# run locally\nnpx nx dev remote-mcp-server\n```\n\nDeberías poder abrir [`http://localhost:8787/`](http://localhost:8787/) en tu navegador\n\n## Conecte el inspector MCP a su servidor\n\nPara explorar su nueva API de MCP, puede utilizar el [Inspector de MCP](https://modelcontextprotocol.io/docs/tools/inspector) .\n\n* Inícielo con `npx @modelcontextprotocol/inspector`\n* [Dentro del inspector](http://localhost:5173) , cambie el Tipo de transporte a `SSE` e ingrese `http://localhost:8787/sse` como la URL del servidor MCP al que conectarse y haga clic en \"Conectar\".\n* Accederá a una pantalla de inicio de sesión (simulada) con usuario y contraseña. Ingrese su correo electrónico y contraseña para iniciar sesión.\n* Deberías ser redirigido nuevamente al Inspector MCP y ahora podrás enumerar y llamar a cualquier herramienta definida.\n\n## Conecte Claude Desktop a su servidor MCP local\n\nEl inspector MCP es genial, pero queremos conectarlo con Claude. Sigue [la guía de inicio rápido de Anthropic](https://modelcontextprotocol.io/quickstart/user) y, en Claude Desktop, ve a Configuración > Desarrollador > Editar configuración para encontrar tu archivo de configuración.\n\nAbra el archivo en su editor de texto y reemplácelo con esta configuración:\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"http://localhost:8787/sse\"\n ]\n }\n }\n}\n```\n\nEsto ejecutará un proxy local y permitirá que Claude se comunique con su servidor MCP a través de HTTP\n\nAl abrir Claude, debería abrirse una ventana del navegador que te permitirá iniciar sesión. Verás las herramientas disponibles en la esquina inferior derecha. Si se le da la indicación correcta, Claude debería solicitar que se ejecute la herramienta.\n\n## Implementar en Cloudflare\n\n1. `npx wrangler kv namespace create OAUTH_KV`\n2. Siga las instrucciones para agregar el ID del espacio de nombres kv a `wrangler.jsonc`\n3. `npm run deploy`\n\n## Llame a su servidor MCP remoto recién implementado desde un cliente MCP remoto\n\nTal como lo hizo anteriormente en \"Desarrollar localmente\", ejecute el inspector MCP:\n\n`npx @modelcontextprotocol/inspector@latest`\n\nLuego ingrese la URL de `workers.dev` (ej: `worker-name.account-name.workers.dev/sse` ) de su Worker en el inspector como la URL del servidor MCP al cual conectarse y haga clic en \"Conectar\".\n\nAhora se ha conectado a su servidor MCP desde un cliente MCP remoto.\n\n## Conecte Claude Desktop a su servidor MCP remoto\n\nActualice el archivo de configuración de Claude para que apunte a la URL de `workers.dev` (por ejemplo: `worker-name.account-name.workers.dev/sse` ) y reinicie Claude\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"https://worker-name.account-name.workers.dev/sse\"\n ]\n }\n }\n}\n```\n\n## Depuración\n\nSi algo sale mal, puede ser útil reiniciar Claude o intentar conectarse directamente a su servidor MCP en la línea de comando con el siguiente comando.\n\n```bash\nnpx mcp-remote http://localhost:8787/sse\n```\n\nEn algunos casos excepcionales puede ser útil borrar los archivos agregados a `~/.mcp-auth`\n\n```bash\nrm -rf ~/.mcp-auth\n```","Spanish","es-ES",{"_29":55,"_33":59,"_38":60},{"_31":56,"_33":57,"_35":58},"Claude などの AI アシスタントが OAuth 認証を介して外部ツールにアクセスできるようにする、Cloudflare Workers ベースのモデルコンテキストプロトコルサーバー。","リモートMCPサーバー","# Cloudflare上のリモートMCPサーバー\n\nOAuth ログインを備えた Cloudflare Workers でリモート MCP サーバーを起動して実行してみましょう。\n\n## ローカルで開発する\n\n```bash\n# clone the repository\ngit clone git@github.com:cloudflare/ai.git\n\n# install dependencies\ncd ai\nnpm install\n\n# run locally\nnpx nx dev remote-mcp-server\n```\n\nブラウザで[`http://localhost:8787/`](http://localhost:8787/)を開くことができるはずです。\n\n## MCPインスペクターをサーバーに接続します\n\n新しい MCP API を調べるには、 [MCP Inspector](https://modelcontextprotocol.io/docs/tools/inspector)を使用できます。\n\n* `npx @modelcontextprotocol/inspector`で起動します。\n* [インスペクタ内で](http://localhost:5173)、トランスポートタイプを`SSE`に切り替え、接続する MCP サーバーの URL として`http://localhost:8787/sse`を入力し、「接続」をクリックします。\n* （模擬）ユーザー名/パスワードログイン画面が表示されます。任意のメールアドレスとパスワードを入力してログインしてください。\n* MCP インスペクターにリダイレクトされ、定義されたツールを一覧表示して呼び出すことができるようになります。\n\n## Claude DesktopをローカルMCPサーバーに接続します\n\nMCP インスペクターは素晴らしいですが、これを Claude に接続したいと思っています。Anthropic[のクイックスタート](https://modelcontextprotocol.io/quickstart/user)に従い、Claude Desktop 内で「設定」>「開発」>「構成の編集」に移動して構成ファイルを見つけてください。\n\nテキストエディターでファイルを開き、次の構成に置き換えます。\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"http://localhost:8787/sse\"\n ]\n }\n }\n}\n```\n\nこれにより、ローカルプロキシが実行され、ClaudeがHTTP経由でMCPサーバーと通信できるようになります。\n\nClaude を開くとブラウザウィンドウが開き、ログインできるようになります。右下に利用可能なツールが表示されます。適切なプロンプトが表示されたら、Claude はツールの呼び出しを要求します。\n\n## Cloudflareにデプロイする\n\n1. `npx wrangler kv namespace create OAUTH_KV`\n2. ガイダンスに従って、kv名前空間IDを`wrangler.jsonc`に追加します。\n3. `npm run deploy`\n\n## リモート MCP クライアントから新しく展開したリモート MCP サーバーを呼び出す\n\n上記の「ローカルで開発する」と同じように、MCP インスペクターを実行します。\n\n`npx @modelcontextprotocol/inspector@latest`\n\n次に、接続先の MCP サーバーの URL として、インスペクターにワーカーの`workers.dev` URL (例: `worker-name.account-name.workers.dev/sse` ) を入力し、「接続」をクリックします。\n\nこれで、リモート MCP クライアントから MCP サーバーに接続されました。\n\n## Claude DesktopをリモートMCPサーバーに接続します\n\nClaude 構成ファイルを更新して、 `workers.dev` URL (例: `worker-name.account-name.workers.dev/sse` ) を参照し、Claude を再起動します。\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"https://worker-name.account-name.workers.dev/sse\"\n ]\n }\n }\n}\n```\n\n## デバッグ\n\n何か問題が発生した場合は、Claude を再起動するか、次のコマンドを使用してコマンドラインで MCP サーバーに直接接続してみると役立つことがあります。\n\n```bash\nnpx mcp-remote http://localhost:8787/sse\n```\n\nまれに、 `~/.mcp-auth`に追加されたファイルを消去すると解決する場合があります。\n\n```bash\nrm -rf ~/.mcp-auth\n```","Japanese","ja-JP",{"_29":62,"_33":66,"_38":67},{"_31":63,"_33":64,"_35":65},"Claude와 같은 AI 도우미가 OAuth 인증을 통해 외부 도구에 액세스할 수 있도록 하는 Cloudflare Workers 기반 모델 컨텍스트 프로토콜 서버입니다.","원격 MCP 서버","# Cloudflare의 원격 MCP 서버\n\nCloudflare Workers에서 OAuth 로그인을 포함한 원격 MCP 서버를 구동해 보겠습니다!\n\n## 지역적으로 개발하다\n\n지엑스피1\n\n브라우저에서 [`http://localhost:8787/`](http://localhost:8787/) 열 수 있어야 합니다.\n\n## MCP 검사기를 서버에 연결하세요\n\n새로운 MCP API를 살펴보려면 [MCP Inspector를](https://modelcontextprotocol.io/docs/tools/inspector) 사용하세요.\n\n* `npx @modelcontextprotocol/inspector` 로 시작하세요\n* [검사기 내에서](http://localhost:5173) 전송 유형을 `SSE` 로 전환하고 연결할 MCP 서버의 URL로 `http://localhost:8787/sse` 입력한 후 \"연결\"을 클릭합니다.\n* (가상) 사용자/비밀번호 로그인 화면으로 이동합니다. 이메일 주소와 비밀번호를 입력하여 로그인하세요.\n* MCP 검사기로 다시 돌아가서 정의된 도구를 나열하고 호출할 수 있습니다!\n\n## Claude Desktop을 로컬 MCP 서버에 연결하세요\n\nMCP 검사기는 훌륭하지만, Claude에 연결하고 싶습니다! [Anthropic의 빠른 시작을](https://modelcontextprotocol.io/quickstart/user) 따라 Claude Desktop에서 설정 > 개발자 > 구성 편집으로 이동하여 구성 파일을 찾으세요.\n\n텍스트 편집기에서 파일을 열고 다음 구성으로 바꾸세요.\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"http://localhost:8787/sse\"\n ]\n }\n }\n}\n```\n\n이렇게 하면 로컬 프록시가 실행되고 Claude가 HTTP를 통해 MCP 서버와 통신할 수 있습니다.\n\nClaude를 열면 브라우저 창이 열리고 로그인할 수 있습니다. 오른쪽 하단에 사용 가능한 도구가 표시됩니다. 오른쪽 프롬프트가 나타나면 Claude가 도구를 호출할지 묻습니다.\n\n## Cloudflare에 배포\n\n1. `npx wrangler kv namespace create OAUTH_KV`\n2. `wrangler.jsonc` 에 kv 네임스페이스 ID를 추가하려면 지침을 따르세요.\n3. `npm run deploy`\n\n## 원격 MCP 클라이언트에서 새로 배포된 원격 MCP 서버를 호출합니다.\n\n위의 \"로컬에서 개발\"에서 한 것처럼 MCP 검사기를 실행합니다.\n\n`npx @modelcontextprotocol/inspector@latest`\n\n그런 다음 검사기에 Worker의 `workers.dev` URL(예: `worker-name.account-name.workers.dev/sse` )을 연결할 MCP 서버의 URL로 입력하고 \"연결\"을 클릭합니다.\n\n이제 원격 MCP 클라이언트에서 MCP 서버에 연결되었습니다.\n\n## Claude Desktop을 원격 MCP 서버에 연결합니다.\n\nClaude 구성 파일을 업데이트하여 `workers.dev` URL(예: `worker-name.account-name.workers.dev/sse` )을 가리키고 Claude를 다시 시작합니다.\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"https://worker-name.account-name.workers.dev/sse\"\n ]\n }\n }\n}\n```\n\n## 디버깅\n\n문제가 발생하면 Claude를 다시 시작하거나 다음 명령을 사용하여 명령줄에서 MCP 서버에 직접 연결을 시도하면 도움이 될 수 있습니다.\n\n```bash\nnpx mcp-remote http://localhost:8787/sse\n```\n\n드물지만 `~/.mcp-auth` 에 추가된 파일을 지우는 것이 도움이 될 수 있습니다.\n\n```bash\nrm -rf ~/.mcp-auth\n```","Korean","ko-KR",{"_29":69,"_33":73,"_38":74},{"_31":70,"_33":71,"_35":72},"Сервер Model Context Protocol на базе Cloudflare Workers, который позволяет помощникам на базе искусственного интеллекта, таким как Клод, получать доступ к внешним инструментам через аутентификацию OAuth.","Удаленный MCP-сервер","# Удаленный сервер MCP на Cloudflare\n\nДавайте запустим удаленный сервер MCP на Cloudflare Workers с возможностью входа по протоколу OAuth!\n\n## Развивайтесь локально\n\n```bash\n# clone the repository\ngit clone git@github.com:cloudflare/ai.git\n\n# install dependencies\ncd ai\nnpm install\n\n# run locally\nnpx nx dev remote-mcp-server\n```\n\nВы должны иметь возможность открыть [`http://localhost:8787/`](http://localhost:8787/) в своем браузере.\n\n## Подключите MCP-инспектор к вашему серверу\n\nДля изучения нового API MCP вы можете использовать [MCP Inspector](https://modelcontextprotocol.io/docs/tools/inspector) .\n\n* Запустите его с помощью `npx @modelcontextprotocol/inspector`\n* [В инспекторе](http://localhost:5173) измените тип транспорта на `SSE` и введите `http://localhost:8787/sse` в качестве URL-адреса сервера MCP для подключения, затем нажмите «Подключиться».\n* Вы перейдете на экран входа в систему (фиктивный) пользователя/пароля. Введите любой адрес электронной почты и пароль для входа.\n* Вас перенаправят обратно в MCP Inspector, и теперь вы сможете просматривать и вызывать любые определенные инструменты!\n\n## Подключите Claude Desktop к локальному серверу MCP\n\nИнспектор MCP великолепен, но мы действительно хотим подключить его к Claude! Следуйте [краткому руководству Anthropic](https://modelcontextprotocol.io/quickstart/user) и в Claude Desktop перейдите в Настройки > Разработчик > Изменить конфигурацию, чтобы найти файл конфигурации.\n\nОткройте файл в текстовом редакторе и замените его следующей конфигурацией:\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"http://localhost:8787/sse\"\n ]\n }\n }\n}\n```\n\nЭто запустит локальный прокси-сервер и позволит Клоду взаимодействовать с вашим сервером MCP по протоколу HTTP.\n\nКогда вы открываете Claude, должно открыться окно браузера, позволяющее вам войти в систему. Вы должны увидеть доступные инструменты в правом нижнем углу. Получив правильный запрос, Claude должен попросить вызвать инструмент.\n\n## Развертывание в Cloudflare\n\n1. `npx wrangler kv namespace create OAUTH_KV`\n2. Следуйте инструкциям, чтобы добавить идентификатор пространства имен kv в `wrangler.jsonc`\n3. `npm run deploy`\n\n## Вызовите недавно развернутый удаленный сервер MCP с удаленного клиента MCP\n\nТак же, как вы делали выше в разделе «Локальная разработка», запустите инспектор MCP:\n\n`npx @modelcontextprotocol/inspector@latest`\n\nЗатем введите URL-адрес `workers.dev` (например: `worker-name.account-name.workers.dev/sse` ) вашего Worker в инспекторе в качестве URL-адреса сервера MCP для подключения и нажмите «Подключиться».\n\nТеперь вы подключились к своему серверу MCP с удаленного клиента MCP.\n\n## Подключите Claude Desktop к удаленному серверу MCP\n\nОбновите файл конфигурации Claude, чтобы он указывал на URL-адрес вашего `workers.dev` (например: `worker-name.account-name.workers.dev/sse` ), и перезапустите Claude.\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"https://worker-name.account-name.workers.dev/sse\"\n ]\n }\n }\n}\n```\n\n## Отладка\n\nЕсли что-то пойдет не так, может быть полезно перезапустить Claude или попробовать подключиться напрямую к вашему серверу MCP через командную строку с помощью следующей команды.\n\n```bash\nnpx mcp-remote http://localhost:8787/sse\n```\n\nВ некоторых редких случаях может помочь очистка файлов, добавленных в `~/.mcp-auth`\n\n```bash\nrm -rf ~/.mcp-auth\n```","Russian","ru-RU",{"_29":76,"_33":80,"_38":81},{"_31":77,"_33":78,"_35":79},"基于 Cloudflare Workers 的模型上下文协议服务器，使 Claude 等 AI 助手能够通过 OAuth 身份验证访问外部工具。","远程 MCP 服务器","# Cloudflare 上的远程 MCP 服务器\n\n让我们在 Cloudflare Workers 上启动并运行一个远程 MCP 服务器，并完成 OAuth 登录！\n\n## 本地开发\n\n```bash\n# clone the repository\ngit clone git@github.com:cloudflare/ai.git\n\n# install dependencies\ncd ai\nnpm install\n\n# run locally\nnpx nx dev remote-mcp-server\n```\n\n您应该能够在浏览器中打开[`http://localhost:8787/`](http://localhost:8787/)\n\n## 将 MCP 检查器连接到您的服务器\n\n要探索新的 MCP api，您可以使用[MCP 检查器](https://modelcontextprotocol.io/docs/tools/inspector)。\n\n* 使用`npx @modelcontextprotocol/inspector`启动\n* [在检查器中](http://localhost:5173)，将传输类型切换为`SSE` ，并输入`http://localhost:8787/sse`作为要连接的 MCP 服务器的 URL，然后点击“连接”\n* 您将导航到（模拟）用户/密码登录屏幕。输入任意电子邮件地址并登录。\n* 您应该被重定向回 MCP 检查器，现在您可以列出并调用任何定义的工具！\n\n## 将 Claude Desktop 连接到您的本地 MCP 服务器\n\nMCP 检查器很棒，但我们更想把它连接到 Claude！按照[Anthropic 的快速入门指南](https://modelcontextprotocol.io/quickstart/user)操作，在 Claude Desktop 中，前往“设置”>“开发者”>“编辑配置”找到你的配置文件。\n\n在文本编辑器中打开该文件并将其替换为以下配置：\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"http://localhost:8787/sse\"\n ]\n }\n }\n}\n```\n\n这将运行本地代理并让 Claude 通过 HTTP 与您的 MCP 服务器通信\n\n打开 Claude 后，会打开一个浏览器窗口并允许您登录。您会在右下角看到可用的工具。输入正确的提示后，Claude 会要求调用该工具。\n\n## 部署到 Cloudflare\n\n1. `npx wrangler kv namespace create OAUTH_KV`\n2. 按照指南将 kv 命名空间 ID 添加到`wrangler.jsonc`\n3. `npm run deploy`\n\n## 从远程 MCP 客户端调用新部署的远程 MCP 服务器\n\n就像上面在“本地开发”中所做的那样，运行 MCP 检查器：\n\n`npx @modelcontextprotocol/inspector@latest`\n\n然后在检查器中输入您的 Worker 的`workers.dev` URL（例如： `worker-name.account-name.workers.dev/sse` ）作为要连接的 MCP 服务器的 URL，然后点击“连接”。\n\n您现在已从远程 MCP 客户端连接到您的 MCP 服务器。\n\n## 将 Claude Desktop 连接到远程 MCP 服务器\n\n更新 Claude 配置文件以指向您的`workers.dev` URL（例如： `worker-name.account-name.workers.dev/sse` ）并重新启动 Claude\n\n```json\n{\n \"mcpServers\": {\n \"math\": {\n \"command\": \"npx\",\n \"args\": [\n \"mcp-remote\",\n \"https://worker-name.account-name.workers.dev/sse\"\n ]\n }\n }\n}\n```\n\n## 调试\n\n如果出现任何问题，重新启动 Claude 或尝试使用以下命令在命令行上直接连接到 MCP 服务器会有所帮助。\n\n```bash\nnpx mcp-remote http://localhost:8787/sse\n```\n\n在极少数情况下，清除添加到`~/.mcp-auth`的文件可能会有所帮助\n\n```bash\nrm -rf ~/.mcp-auth\n```","Chinese","zh-CN","githubUser","locale",{"_85":86,"_87":88,"_13":95,"_98":42,"_99":42,"_100":101,"_104":105,"_125":-5,"_16":126,"_127":128,"_129":130,"_33":43,"_148":118,"_149":150,"_153":-5,"_154":-5,"_155":-5,"_156":157,"_161":-5,"_137":114,"_162":-5,"_163":-5,"_119":164,"_165":166},"addedAt",["D",1746669721468],"argumentsJsonSchema",{"_89":90,"_91":92,"_93":94},"properties",{},"required",[],"type","object",[96,97],"language:typescript","license_grade:f","descriptionMarkdown","descriptionPlainText","githubProject",{"_35":44,"_102":103},"url","https://github.com/cyyzpanpan/remote-mcp-server","githubRepository",{"_106":107,"_108":109,"_110":111,"_112":113,"_33":114,"_115":116,"_121":122,"_123":-5,"_124":122},"defaultBranch","main","fullName","cyyzpanpan/remote-mcp-server","githubTreeId",116102,"language","TypeScript","remote-mcp-server","owner",{"_117":118,"_119":120},"login","cyyzpanpan","uid","9IdxJIdizR","recentStargazers",0,"spdxLicense","stargazers","glamaJson",6435,"inspectable",false,"integrations",[131,141],{"_132":133,"_31":139,"_119":140},"brand",{"_134":135,"_33":136,"_137":138},"logoSvgPath","M16.5088 16.8447c.1475-.5068.0908-.9707-.1553-1.3154-.2246-.3164-.6045-.499-1.0615-.5205l-8.6592-.1123a.1559.1559 0 0 1-.1333-.0713c-.0283-.042-.0351-.0986-.021-.1553.0278-.084.1123-.1484.2036-.1562l8.7359-.1123c1.0351-.0489 2.1601-.8868 2.5537-1.9136l.499-1.3013c.0215-.0561.0293-.1128.0147-.168-.5625-2.5463-2.835-4.4453-5.5499-4.4453-2.5039 0-4.6284 1.6177-5.3876 3.8614-.4927-.3658-1.1187-.5625-1.794-.499-1.2026.119-2.1665 1.083-2.2861 2.2856-.0283.31-.0069.6128.0635.894C1.5683 13.171 0 14.7754 0 16.752c0 .1748.0142.3515.0352.5273.0141.083.0844.1475.1689.1475h15.9814c.0909 0 .1758-.0645.2032-.1553l.12-.4268zm2.7568-5.5634c-.0771 0-.1611 0-.2383.0112-.0566 0-.1054.0415-.127.0976l-.3378 1.1744c-.1475.5068-.0918.9707.1543 1.3164.2256.3164.6055.498 1.0625.5195l1.8437.1133c.0557 0 .1055.0263.1329.0703.0283.043.0351.1074.0214.1562-.0283.084-.1132.1485-.204.1553l-1.921.1123c-1.041.0488-2.1582.8867-2.5527 1.914l-.1406.3585c-.0283.0713.0215.1416.0986.1416h6.5977c.0771 0 .1474-.0489.169-.126.1122-.4082.1757-.837.1757-1.2803 0-2.6025-2.125-4.727-4.7344-4.727","Cloudflare","slug","cloudflare","Allows deploying the MCP server to Cloudflare's infrastructure with OAuth login capabilities.","0xtmc4k21u",{"_132":142,"_31":146,"_119":147},{"_134":143,"_33":144,"_137":145},"m8.213.063 8.879 12.136-8.67 11.739h2.476l8.665-11.735-8.89-12.14Zm4.728 0 9.02 11.992-9.018 11.883h2.496L24 12.656v-1.199L15.434.063ZM7.178 2.02.01 11.398l-.01 1.2 7.203 9.644 1.238-1.676-6.396-8.556 6.361-8.313Z","Cloudflare Workers","cloudflare-workers","Provides serverless hosting for the MCP server, allowing it to run in Cloudflare's edge network with KV storage for OAuth functionality.","lj1mgvylzq","namespace","npmPackage",{"_33":114,"_151":152},"weeklyDownloads",14,"prompts","redditPostForeignId","resources","scores",{"_158":122,"_159":-5,"_160":-5},"license","quality","security","simpleIcon","supportedPlatforms","tools","2u5b9z6ate","updatedAt",["D",1746669733175],"meta",[169,172,174,181,184,186,189,192,195,198,201,203,206,209,211,213,215,217,219],{"_170":171},"title","Remote MCP Server | Glama",{"_173":42,"_33":31},"content",{"_175":176,"_177":178,"_179":180},"href","https://glama.ai/mcp/servers/@cyyzpanpan/remote-mcp-server","rel","canonical","tagName","link",{"_173":43,"_182":183},"property","og:title",{"_173":42,"_182":185},"og:description",{"_173":187,"_182":188},"https://glama.ai/logo.png","og:logo",{"_173":190,"_182":191},"https://glama.ai/mcp/servers/@cyyzpanpan/remote-mcp-server/og-image","og:image",{"_173":193,"_182":194},"630","og:image:height",{"_173":196,"_182":197},"1200","og:image:width",{"_173":199,"_182":200},"website","og:type",{"_173":176,"_182":202},"og:url",{"_173":204,"_182":205},"Glama – MCP Hosting Platform","og:site_name",{"_173":207,"_33":208},"summary_large_image","twitter:card",{"_173":204,"_33":210},"twitter:title",{"_173":42,"_33":212},"twitter:description",{"_173":190,"_33":214},"twitter:image",{"_173":193,"_33":216},"twitter:image:height",{"_173":196,"_33":218},"twitter:image:width",{"_220":221},"script:ld+json",{"_222":223,"_224":225},"@context","http://schema.org","@graph",[226],{"_222":227,"_228":229,"_230":231},"https://schema.org","@type","BreadcrumbList","itemListElement",[232,239,243,247],{"_228":233,"_234":235,"_33":236,"_237":238},"ListItem","item","https://glama.ai/","Glama","position",1,{"_228":233,"_234":240,"_33":241,"_237":242},"https://glama.ai/mcp","MCP",2,{"_228":233,"_234":244,"_33":245,"_237":246},"https://glama.ai/mcp/servers","Servers",3,{"_228":233,"_234":176,"_33":43,"_237":248},4,"visitorIsMaintainer","routes/_public/mcp/servers/~namespace/~slug/_pages/reviews/_route",{"_82":-5,"_29":252,"_167":269,"_304":-5},{"_85":253,"_87":254,"_13":257,"_98":42,"_99":42,"_100":258,"_104":259,"_125":-5,"_16":126,"_127":128,"_129":261,"_33":43,"_148":118,"_149":266,"_153":-5,"_154":-5,"_155":-5,"_156":267,"_161":-5,"_137":114,"_162":-5,"_163":-5,"_119":164,"_165":268},["D",1746669721468],{"_89":255,"_91":256,"_93":94},{},[],[96,97],{"_35":44,"_102":103},{"_106":107,"_108":109,"_110":111,"_112":113,"_33":114,"_115":260,"_121":122,"_123":-5,"_124":122},{"_117":118,"_119":120},[262,264],{"_132":263,"_31":139,"_119":140},{"_134":135,"_33":136,"_137":138},{"_132":265,"_31":146,"_119":147},{"_134":143,"_33":144,"_137":145},{"_33":114,"_151":152},{"_158":122,"_159":-5,"_160":-5},["D",1746669733175],[270,272,274,276,278,279,280,281,282,283,284,285,286,287,288,289,290,291,292],{"_170":271},"Reviews | Remote MCP Server | Glama",{"_173":273,"_33":31},"Read community reviews and feedback for Remote MCP Server. Discover user experiences, ratings, strengths, and potential limitations of this MCP implementation from actual users.",{"_175":275,"_177":178,"_179":180},"https://glama.ai/mcp/servers/@cyyzpanpan/remote-mcp-server/reviews",{"_173":277,"_182":183},"Reviews | Remote MCP Server",{"_173":273,"_182":185},{"_173":187,"_182":188},{"_173":190,"_182":191},{"_173":193,"_182":194},{"_173":196,"_182":197},{"_173":199,"_182":200},{"_173":275,"_182":202},{"_173":204,"_182":205},{"_173":207,"_33":208},{"_173":204,"_33":210},{"_173":273,"_33":212},{"_173":190,"_33":214},{"_173":193,"_33":216},{"_173":196,"_33":218},{"_220":293},{"_222":223,"_224":294},[295],{"_222":227,"_228":229,"_230":296},[297,298,299,300,301],{"_228":233,"_234":235,"_33":236,"_237":238},{"_228":233,"_234":240,"_33":241,"_237":242},{"_228":233,"_234":244,"_33":245,"_237":246},{"_228":233,"_234":176,"_33":43,"_237":248},{"_228":233,"_234":275,"_33":302,"_237":303},"Reviews",5,"review","actionData","errors"]

Rate & Review