Render MCP Server – Services and Deployments

Overview

Render's official MCP server connects compatible AI applications to Render
infrastructure through natural-language tools. It lets an agent work with live
workspaces, services, deploys, logs, metrics, and datastores instead of relying
only on manually copied dashboard information or static model knowledge.

What the MCP server enables

Depending on the selected workspace and the permissions of the authenticated
Render account, an AI agent can:

• List accessible workspaces and select the active workspace.
• Create web services, static sites, cron jobs, Render Postgres databases, and
  Render Key Value instances.
• List services and retrieve service details.
• Replace a service's environment variables.
• List deploy history and retrieve details for a specific deploy.
• Filter service logs and inspect available log-label values.
• Retrieve CPU, memory, instance-count, connection, response, latency, and
  outbound-bandwidth metrics where supported by the workspace plan.
• List Render Postgres databases and run read-only SQL queries.
• List and inspect Render Key Value instances.

When to use it

Use Render MCP when an AI workflow needs current infrastructure context or must
perform an approved Render operation. Typical examples include creating a new
web service, provisioning a database, checking recent deploys, analyzing
service logs, investigating traffic or autoscaling behavior, querying a
read-only database replica, and reviewing datastore health from an AI coding
assistant.

Connection and authentication

Render hosts the recommended Streamable HTTP endpoint at
https://mcp.render.com/mcp. Clients authenticate with a Render API key in the
Authorization header as a Bearer token.

Render also publishes an official Docker image for local stdio use at
ghcr.io/render-oss/render-mcp-server. The local server reads
RENDER_API_KEY from the environment. Render strongly recommends the hosted
endpoint because it automatically receives new capabilities.

Key considerations

Render API keys are broadly scoped and grant access to every workspace and
service available to the account. Use a dedicated account or key where
appropriate, store credentials in a protected secret manager, and connect only
trusted MCP clients. The server currently supports a limited set of creation
options and does not support most modifications, deletions, manual deploy
triggers, or scaling changes. Updating environment variables is the main
supported modification to an existing service. Render also warns that sensitive
information might still be exposed to the AI client, so review prompts and
tool results carefully.