Render MCP Server

The Render MCP Server is the official Model Context Protocol server from Render, Inc. for the eponymous cloud platform. It connects compatible AI clients — including Claude Code, Claude Desktop, Codex Desktop, Codex CLI, and Cursor — directly to the Render dashboard and Render API. Developers can manage cloud infrastructure in natural language without switching between terminal, browser, and code editor. The direct product link to the official documentation is https://render.com/docs/mcp-server; the open-source repository lives at https://github.com/render-oss/render-mcp-server.

What is the Render MCP Server?

According to the provider, Render is a cloud platform for web services, static sites, databases, cron jobs, and AI infrastructure. The MCP server extends this platform with a standardized interface for AI agents. It uses the Model Context Protocol (MCP), an open standard for connecting AI applications to external tools and data sources. When an agent receives a Render-relevant prompt, the MCP client delegates the task to the server, which in turn calls the Render API and returns the result in a structured format. The server is hosted and reachable at https://mcp.render.com/mcp — no local installation step is required for basic operation.

Prerequisites

Using the Render MCP Server requires: an active Render account, a compatible MCP client (Claude Code, Claude Desktop, Codex, or Cursor), and either OAuth consent or a Render API key for authentication. For API key setups, Node.js should be installed so that the mcp-remote package can be used. No local build step is needed because the server runs remotely at https://mcp.render.com/mcp. The source code repository is Apache-2.0 licensed for custom modifications or self-hosting.

Features and tools

The Render MCP Server provides several tool groups according to the README and official documentation:

Workspaces: list_workspaces lists accessible workspaces, select_workspace selects a workspace, and get_selected_workspace returns the current context. Every resource tool also accepts an optional workspaceId parameter for explicit assignment across session boundaries.

Services: list_services lists all services in the Render account, and get_service retrieves details for a single service. create_web_service creates new web services — including runtime selection (Node, Python, Go, Rust, Ruby, Elixir, Docker), build and start commands, region, auto-deploy, and environment variables. Analogous tools are create_static_site for static sites and create_cron_job for scheduled tasks. update_environment_variables updates environment variables; direct service updates are deliberately not supported and instead point to the dashboard or Render API.

Deployments: list_deploys shows the deployment history of a service, get_deploy fetches details for a specific deploy, and trigger_deploy triggers a new deployment — optionally clearing build caches.

Logs: list_logs filters logs by resource, severity level, type, instance, host, status codes, HTTP methods, and time windows. list_log_label_values lists all possible values for a given log label for targeted analysis.

Metrics: get_metrics retrieves performance metrics for services, Postgres databases, and key-value stores. Supported metrics include CPU usage, memory usage, instance count, HTTP request counts, HTTP latency, bandwidth, and active connections.

Databases: PostgreSQL resources can be managed via list_postgres, get_postgres, create_postgres, list_postgres_databases, create_postgres_database, list_postgres_roles, and get_connection_info. The execute_query tool allows Postgres databases to be queried directly from the AI client. Key-value stores are accessible through list_key_value_stores, get_key_value_store, and create_key_value_store.

Projects: list_projects, get_project, and create_project enable project management in the Render dashboard.

Setup

Setup varies by client. Claude Code supports an official Render plugin (/plugin install render@claude-plugins-official) for OAuth and automatic skill integration. Alternatively, the server can be configured with an API key directly: claude mcp add --transport http render https://mcp.render.com/mcp --header "Authorization: Bearer <API_KEY>". Claude Desktop uses the Connectors interface for OAuth or adds the mcp-remote call to claude_desktop_config.json. Codex uses codex mcp add render --url https://mcp.render.com/mcp --oauth-client-id codex for OAuth or a TOML configuration for API keys. Cursor supports the Render plugin via /add-plugin render or a JSON configuration at ~/.cursor/mcp.json. API keys are created in the Render dashboard under Account Settings.

Security

The Render MCP Server supports potentially destructive operations according to the provider: modifying environment variables, triggering new deployments, and creating new resources. Before setup, consider which permissions are granted to the API key or OAuth consent. API keys belong in environment variables or secure client configurations — not in publicly visible files or prompts. Users relying on OAuth should regularly review the scopes granted in the Render dashboard. Prompt injection via manipulated log content or service names is a known risk for MCP servers with network and database access; agent outputs should therefore be reviewed before any production changes.

FAQ

Is the Render MCP Server free? The server itself is Apache-2.0 licensed and open source. Usage depends on the costs of the respective Render account and the AI client used. Render publishes current pricing information at https://render.com/pricing.

Is OAuth mandatory? No. OAuth is the recommended path for interactive clients like Claude Desktop and Cursor, but all listed clients also support API key authentication for non-interactive or CI/CD environments.

Can I self-host the server? Yes. The Apache-2.0 repository at https://github.com/render-oss/render-mcp-server contains the source code for custom deployments or modifications.

How current are the GitHub stars? The GitHub API reported exactly 165 stars on 2026-09-08. That number is a point-in-time snapshot of repository popularity and is not a security or quality promise.

Render account, MCP client (Claude Code, Claude Desktop, Codex, or Cursor), and OAuth consent or a Render API key.

For Claude Code: /plugin install render@claude-plugins-official for OAuth or claude mcp add --transport http render https://mcp.render.com/mcp --header "Authorization: Bearer <API_KEY>" for API key setup. For other clients, follow the respective plugin or configuration documentation.

OAuth (recommended for interactive clients) or Render API key for non-interactive and CI/CD environments.

Permissions follow the Render account, OAuth scopes, or the permission scope of the API key; destructive operations are possible.

Tool calls reach the Render API and return structured results to the MCP client; data does not leave the Render platform additionally.

Destructive operations (modifying environment variables, triggering deployments), API keys in insecure configurations, and prompt injection via log or service names.