list_workspacesLABSCO SUMMARY
Render's own server, so deploying and debugging are one conversation.
Creating the service, watching the deploy, reading the logs when it fails and querying the database behind it are usually four places; here they are one session, with the workspace named on each call so the account being touched is explicit. The read side is the stronger half — logs with real filters, metrics with quantiles and aggregation — while configuration changes past environment variables still route back to the dashboard.
What it is
Render's own server for the resources in your account: web services, static sites, cron jobs, Postgres databases and key-value stores, plus deploys, logs and metrics for all of them.
What you get
- Services listed and read in detail, with previews included when you ask for them, and the workspace passed explicitly on each call so which account is being changed is never a guess
- A web service created with its runtime — node, python, go, rust, ruby, elixir or docker — plus build and start commands, repo, branch, plan and region; static sites and cron jobs created the same way
- Environment variables replaced as a set on an existing service
- Deploy history per service, one deploy read in detail, and a deploy triggered by hand — with the build cache cleared if you ask
- Logs filtered the way an incident actually needs: by resource, severity, type, instance, host, status code, method, path, text and time window, plus the values a given log label takes
- Metrics for any resource — CPU, memory, instance count, bandwidth, HTTP request counts and latency at a quantile you choose, active connections for databases — over a window up to 30 days back
- Postgres created, listed, described and queried with read-only SQL, and key-value instances created with their eviction and persistence policies
Requirements
A Render account, connected as described in Render's own setup documentation. Creating a web service takes a runtime, a build command and a start command; the region defaults to oregon and auto-deploy to yes. One thing to know before planning around it: the update tools for a service, static site or cron job do not change anything — they hand back a link to that resource's settings in the dashboard, so edits beyond environment variables happen there or through the REST API. Apache-2.0.
Setup effort
Paste a URL, then authorize — add the endpoint to your client, then approve the OAuth consent screen
USE CASES
Deploying something new
Pick the workspace, create the service with its runtime, build and start commands, and let auto-deploy take it from the branch you named.
Working out why a deploy failed
Read the deploy history for the service, open the failing one, then pull the logs for that resource filtered to errors in the window it broke.
Answering "is it the app or the database"
Pull CPU, memory and HTTP latency for the service, then active connections for the database behind it, over the same window.
Checking data without opening a SQL client
Run a read-only query against the Render-hosted Postgres from the same conversation where the question came up.
Redeploying something that will not redeploy itself
Trigger the deploy by hand — clearing the build cache when a stale dependency is the suspect — and watch it through the deploy record.
Scheduling recurring work
Create the cron job with its schedule expression, runtime and commands, in the region and plan you want it to run in.
25 tools
Listed in project docs. Your client may show a different list.
select_workspaceget_selected_workspacelist_servicesget_servicecreate_web_servicecreate_static_sitecreate_cron_jobupdate_environment_variablesupdate_web_serviceupdate_static_siteupdate_cron_joblist_deploysget_deploytrigger_deploylist_logslist_log_label_valuesget_metricsquery_render_postgreslist_postgres_instancesget_postgrescreate_postgreslist_key_valueget_key_valuecreate_key_valueBEFORE YOU INSTALL
What to know first
Tools you get
25 tools named in the project's documentation.
INSTALLATION
Connect your client
Add the endpoint to your client, then approve the OAuth consent screen.
Server setup
https://mcp.render.com/mcpRequirements
- OAuth (official plugins / connector / codex --oauth-client-id)
- API key as Bearer header (created from Account Settings)
MCP Clients
Claude Code — API Key
Create an API key from your Account Settings page, then run the following command; <YOUR_API_KEY> is that key:
claude mcp add --transport http render https://mcp.render.com/mcp --header "Authorization: Bearer <YOUR_API_KEY>"Claude Code — Render Plugin
In Claude Code, run the following command:
/plugin install render@claude-plugins-officialSelect a scope of access for the plugin.
Run /reload-plugins.
The first time Claude uses the MCP server, it opens an authorization flow in your browser to connect your Render account.
Cursor — API Key
Create an API key from your Account Settings page, then add the following configuration to ~/.cursor/mcp.json:
{
"mcpServers": {
"render": {
"url": "https://mcp.render.com/mcp",
"headers": {
"Authorization": "Bearer <YOUR_API_KEY>"
}
}
}
}Codex — OAuth
Run the following command:
codex mcp add render --url https://mcp.render.com/mcp --oauth-client-id codexComplete the authorization flow in your browser to connect your Render account.