-
Notifications
You must be signed in to change notification settings - Fork 1
en MCP Server Management
A Server is the core unit in Arthur MCP: it wraps one data source — today a REST API, with GraphQL, gRPC, and several SQL/NoSQL connectors already visible in the UI — and gets its own MCP endpoint that AI clients can connect to.
Open the Arthur screen that visually supports “MCP Server Management”.

- Create a server from an OpenAPI/Swagger spec, a Postman collection, a REST template, or from a blank slate.
- Configure its base URL, upstream authentication, rate limits, response size limits, availability windows, and maintenance mode.
- Pause a server to stop all traffic instantly without losing any of its configuration, then reactivate it whenever you're ready.
- Re-import an updated API spec into an existing server instead of rebuilding its tools from zero — new and changed operations are picked up without discarding your edits to the rest.
Every server has a slug — a short, unique, URL-safe identifier (lowercase letters, numbers, and hyphens only) auto-generated from the server's name when you create it. It's the server's public address: it's what appears in the public share link and in the MCP endpoint URL AI clients connect to.
You can change it yourself from the Connect tab, under "Public doc slug." Since it must stay unique across every server on the instance, changing it is also how you revoke an old public share link — the previous slug simply stops resolving. See How to Manage Your Server Slug.
The Activity tab is a live execution log for the server: every tool call, with timestamp, tool name, whether it came from an MCP client or a direct test, status code, response time, and any error message. Rows with a request or response body can be expanded to see the full JSON. Use it to debug a failing tool call or to confirm an AI client is actually reaching your server.
Bring the Servers area into the viewport for a focused, semantic screenshot.

- From the Dashboard, click New Server.
- Choose how to start it: upload a spec/collection, pick a template, or start blank. See Getting Started for a walkthrough of each path.
- Open the server and use its tabs — Connect, Tools, Resources, Prompts, Settings, Activity — to shape and operate it.
- Use Settings to adjust the base URL, authentication, and traffic controls at any time, or to pause the server temporarily.
- Pausing a server is reversible and does not delete anything — it simply stops the server from answering MCP calls until reactivated.
- Related: REST API Template Gallery, Dynamic Tools, Settings.
Arthur MCP Wiki
Home · Getting Started · FAQ
⚙️ Features
Servers & Tools
Security & Access
Platform
Operations
Coming Soon
📚 Guides (46)
Connect AI Clients
Security & Access
- Configure API Credentials
- Configure Secrets
- Create a Custom Role
- Generate an Access Key
- Protect a Server with an Access Key
- Rotate the JWT Signing Secret
- Share a Server Publicly (and Revoke Access)
- Store and Reference a Secret
Harness & Limits
- Configure Execution Hooks
- Configure Input Constraints
- Configure Multi-Tenant Parameters
- Configure Output Filtering
- Configure Request Limit
- Configure Response Limits
- Configure Retry Policy
- Configure Timeout Settings
- Configure Tool Restrictions
- Set Global Request Headers
Platform & Operations
- Configure AI Providers
- Connect an Error Tracking Provider
- Enable Error Alerts
- Import the Grafana Dashboard
- Set Up Error Alerts
Tools, Resources & Prompts
- Add HTML Output to a Tool
- Add a Resource
- Build a Chain
- Create Prompts
- Create a Resource
- Create a Tool
- Create an HTML Template in a Resource
- Create and Link a Prompt
- Return an HTML Template from a Tool
- Test a Tool
- Use AI-Assisted Tool Generation
Servers & Import
🔌 API Integrations (230)
AI (28)
- AI/ML API
- Anthropic (Claude)
- Chatwith
- Clarifai
- Cohere
- CustomGPT.ai
- Dialogflow
- Eden
- ElevenLabs
- Gladia
- Groq
- HOL Registry Broker
- Hugging Face
- Imagga
- Irisnet
- MessengerX.io
- Mistral AI
- NLP Cloud
- OpenAI
- Perplexity AI
- Perspective
- Replicate
- Roboflow Universe
- Stability AI
- Summarize Text with AI
- Together AI
- Unplugg
- WolframAlpha
Animals (19)
Anime (14)
Anti-Malware (10)
Art & Design (23)
Books (16)
Business (27)
Calendar (14)
Cloud Storage & File Sharing (10)
Communication (9)
Continuous Integration (5)
Data (15)
Database (3)
Development (8)
E-commerce (5)
Government & Legal (1)
Music & Media (5)
Testing (2)
✨ Prompt Templates (90)
Analysis (6)
Code (9)
Customer Support (6)
Data Extraction (6)
DevOps (5)
Education (5)
Finance (5)
HR & Recruiting (5)
Legal (4)
Product Management (5)
Research (5)
Sales (5)
SEO & Marketing (5)
Social Media (5)
Summarization (6)
💡 Concepts (37)
Connect AI Clients
Security & Access
- Do I Have Swagger Documentation for My MCP Routes?
- How Do Access Keys Work?
- Should I Have a Separate Access Key per AI Client?
- What Are Access Keys For?
- What Are Secrets For?
- What Is API Credentials For?
- What Is API Credentials?
- What Is the "Share the MCP Swagger Documentation" Feature?
- What Do I Do with the QR Code on the Share Page?
Harness & Limits
- What Are Multi-Tenant Parameters For?
- What Are Response Limits For?
- What Is Execution Hooks For?
- What Is Execution Hooks?
- What Is Input Constraints For?
- What Is Input Constraints?
- What Is Output Filtering For?
- What Is Output Filtering?
- What Is Request Limit For?
- What Is Request Limit?
- What Is Retry Policy For?
- What Is Retry Policy?
- What Is Timeout Settings For?
- What Is Timeout Settings?
- What Is Tool Restrictions For?
- What Is Tool Restrictions?
Platform & Operations
Tools, Resources & Prompts
Servers & Import
Início · Primeiros Passos · FAQ
⚙️ Funcionalidades
Servidores & Ferramentas
- Gerenciamento de Servidores MCP
- Galeria de Templates de API REST
- Ferramentas Dinâmicas
- Cadeias
- Recursos
- Prompts
Segurança & Acesso
Plataforma
Operações
Em Breve
📚 Guias (46)
Conectar Clientes de IA
Segurança & Acesso
- Armazenar e Referenciar um Segredo
- Configurar Credenciais da API
- Configurar Secrets?
- Criar um Papel Personalizado
- Gerar uma Access Key?
- Rotacionar o Segredo de Assinatura JWT
Harness & Limites
- Configurar Execution Hooks?
- Configurar Input Constraints?
- Configurar Limites de Resposta
- Configurar Output Filtering?
- Configurar Parâmetros Multi-Tenant
- Configurar Request Limit?
- Configurar Retry Policy?
- Configurar Timeout Settings?
- Configurar Tool Restrictions?
- Definir Cabeçalhos Globais de Requisição
Plataforma & Operações
- Conectar um Provedor de Rastreamento de Erros
- Configurar AI Providers?
- Configurar Alertas de Erro
- Habilitar os Error Alerts?
- Importar o Dashboard do Grafana
Ferramentas, Recursos & Prompts
- Adicionar Saída HTML a uma Ferramenta
- Adicionar um Recurso
- Construir uma Cadeia
- Criar Meus Prompts?
- Criar e Vincular um Prompt
- Criar um Resource?
- Eu Crio um Template de HTML em um Resource?
- Criar uma Tool?
- Retornar um Template HTML em uma Tool?
- Testar uma Ferramenta
- Usar Geração de Ferramentas Assistida por IA
Servidores & Importação
- Compartilhar um Servidor Publicamente (e Revogar o Acesso)
- Deixar Meu Servidor Disponível Apenas em Alguns Dias e Horários?
- Criar um Servidor a partir de um Template
- Criar um Slug
- Gerenciar o Slug do Seu Servidor
- Importar uma Coleção do Postman
- Importar uma Especificação OpenAPI
- Pausar e Reativar um Servidor
- Proteger um Servidor com uma Chave de Acesso
- Reimportar uma Especificação de API Atualizada
🔌 Integrações de API (230)
AI (28)
- AI/ML API
- Anthropic (Claude)
- Chatwith
- Clarifai
- Cohere
- CustomGPT.ai
- Dialogflow
- Eden
- ElevenLabs
- Gladia
- Groq
- HOL Registry Broker
- Hugging Face
- Imagga
- Irisnet
- MessengerX.io
- Mistral AI
- NLP Cloud
- OpenAI
- Perplexity AI
- Perspective
- Replicate
- Roboflow Universe
- Stability AI
- Summarize Text with AI
- Together AI
- Unplugg
- WolframAlpha
Animals (19)
Anime (14)
Anti-Malware (10)
Art & Design (23)
Books (16)
Business (27)
Calendar (14)
Cloud Storage & File Sharing (10)
Communication (9)
Continuous Integration (5)
Data (15)
Database (3)
Development (8)
E-commerce (5)
Government & Legal (1)
Music & Media (5)
Testing (2)
✨ Templates de Prompt (90)
Analysis (6)
Code (9)
Customer Support (6)
Data Extraction (6)
DevOps (5)
Education (5)
Finance (5)
HR & Recruiting (5)
Legal (4)
Product Management (5)
Research (5)
Sales (5)
SEO & Marketing (5)
Social Media (5)
Summarization (6)
💡 Conceitos (37)
Conectar Clientes de IA
Segurança & Acesso
- Como Funcionam as Access Keys?
- Tenho uma Documentação de Swagger para as Minhas Rotas MCP?
- O que É API Credentials?
- O que É a Funcionalidade "Share the MCP Swagger Documentation"?
- Para que Serve API Credentials?
- O que Eu Faço com o QR Code do Share the MCP Swagger Documentation?
- Para que Servem as Access Keys?
- Para que Servem as Secrets?
- Devo Ter uma Access Key para Cada Cliente de IA que Eu Integrar?
Harness & Limites
- O que É Execution Hooks?
- O que É Input Constraints?
- O que É Output Filtering?
- O que É Request Limit?
- O que É Retry Policy?
- O que É Timeout Settings?
- O que É Tool Restrictions?
- Para que Serve Execution Hooks?
- Para que Serve Input Constraints?
- Para que Serve Output Filtering?
- Para que Serve Retry Policy?
- Para que Serve Timeout Settings?
- Para que Serve Tool Restrictions?
- Para que Serve o Request Limit?
- Para que Servem os Multi-Tenant Parameters?
- Para que Servem os Response Limits?
Plataforma & Operações
- Para que Serve a Aba de Activity Log?
- Para que Servem AI Providers?
- Para que Servem os Error Alerts?
Ferramentas, Recursos & Prompts
Servidores & Importação