How is this different from connecting Render MCP directly?

Connected directly, an agent sees every tool the server exposes and holds the credential itself. Through TrustGate you choose which tools each agent gets, and every call is logged against the agent that made it.

What if the Render MCP API changes?

We track the Render MCP schema, so the tool catalogue stays current. Tools you have already allowed keep working, and anything new stays denied until you enable it.

Can I give different agents different Render tools?

Yes. Access is set per agent: one can get a handful of read tools, another the full set. You grant the whole MCP or single tools within it.

Where can I see what my agents did in Render?

Every call is logged against the agent that made it, with the tool it used. Activity can be sent on to your own monitoring.

Do I need to run the Render MCP server myself?

No. TrustGate connects to the Render MCP server and sits in front of it. Your agent only talks to your TrustGate URL, never to Render directly.

Connect Render MCP to your AI agents, securely.

Q: Is Render MCP secure?

The server itself exposes 26 tools, and some of them can write to your Render account. Behind TrustGate an agent only sees the tools you allow, requests and responses are inspected, and every call is logged. Anything outside the allowlist is denied before it reaches Render.

One secured endpoint for every Render call your AI agents make, scoped per user, inspected, and logged.

Get started for free

Get a demo

Try Render MCP now

What your agents can ask once Render MCP is connected.

Select the tools your AI agents can call

26 tools from the Render MCP schema.

Showing 26 tools

list_workspacesreadLists the workspaces the caller has access to. No parameters.
get_selected_workspacereadGets the currently selected workspace. No parameters.
select_workspacewriteSelects a workspace for the session. Deprecated: new code should pass workspaceId on each resource tool call instead, and the implicit session fallback is scheduled for removal.
list_servicesreadLists all services in the account, optionally including preview services.
get_servicereadGets the details of a specific service by ID.
update_web_servicereadDoes not update anything. Returns a link to the service's settings in the Render Dashboard, because direct updates are not supported through the MCP server.
update_static_sitereadDoes not update anything. Returns a link to the site's settings in the Render Dashboard.
update_cron_jobreadDoes not update anything. Returns a link to the cron job's settings in the Render Dashboard.
create_web_servicehigh riskCreates a billable web service with a runtime, build and start commands, repo and branch, plan, auto-deploy setting, region and environment variables.
create_static_sitehigh riskCreates a static site with a build command, repo, branch, publish path, auto-deploy setting and environment variables.
create_cron_jobhigh riskCreates a billable cron job with a cron schedule, runtime, build and start commands, plan, region and environment variables.
update_environment_variableshigh riskUpdates all environment variables for a service. Takes the complete list and replaces the existing set, so omitted variables are removed.
list_deploysreadLists the deployment history for a service.
get_deployreadGets the details of a specific deployment.
trigger_deployhigh riskTriggers a new deploy for a service, optionally clearing the build cache first. Intended for deploys that would not happen automatically.
list_eventsreadList a service's event history: deploys, builds, restarts, failures, scaling, suspensions, and disk changes.
list_logsreadLists logs filtered by resource, severity level, type, instance, host, status code, method, path, text and time range, with direction and limit.
list_log_label_valuesreadLists all values for a given log label across the logs matching the provided filters.
get_metricsreadGets performance metrics for any Render resource: CPU and memory usage, limits and autoscaling targets, instance count, HTTP request count and latency quantiles, bandwidth and active connections. Limited to the last 30 days.
list_postgres_instancesreadLists all PostgreSQL databases in the account. No parameters.
get_postgresreadGets the details of a specific PostgreSQL database.
query_render_postgresreadRuns a read-only SQL query against a Render-hosted Postgres database.
create_postgreshigh riskCreates a billable PostgreSQL database with a plan ranging from free to 1 TB accelerated, plus region, version and disk size.
list_key_valuereadLists all key-value instances in the account. No parameters.
get_key_valuereadGets the details of a specific key-value instance.
create_key_valuehigh riskCreates a billable key-value instance with a plan, region, eviction policy and persistence mode, which can be set to off.

What Render MCP exposes

Render MCP groups its 26 tools into 7 categories, each with a suggested starting policy.

All 26 tools as JSON, ready to paste.

Category	Covers	Access	Suggested policy
Services	Listing and getting services, creating web services, static sites and cron jobs with runtime, build and start commands, plan, region and environment variables, replacing a service's environment variables, and the three update_ tools that return a Dashboard link	Read / Write	update_environment_variables takes the complete list and replaces what is there, so a partial list silently drops variables. The three update_ tools are safe: they only return a link
Postgres	Listing Postgres instances, getting one's details, creating a database with plan, region, version and disk size, and running a read-only SQL query against it	Read / Write	query_render_postgres is read-only but reaches production data directly; create_postgres starts a billable database on plans up to 1 TB
Workspaces	Listing accessible workspaces, selecting one and reading the current selection	Read / Write	select_workspace is deprecated. Pass workspaceId on each resource call instead, so the target stays explicit across reconnects
Deployments	Listing a service's deployment history, getting a specific deployment and triggering a new deploy with optional cache clearing	Read / Write	trigger_deploy is for deploys that would not happen automatically; services with autoDeploy already deploy on push
Key Value	Listing key-value instances, getting one's details and creating one with plan, region, eviction policy and persistence mode	Read / Write	create_key_value starts a billable instance; the persistence mode can be set to off, which prioritises write speed over durability
Logs	Listing logs filtered by resource, level, type, instance, host, status code, method, path, text and time range, and listing the values of a given log label	Read	Read-only, but application logs routinely carry customer data; restrict and cap the limit
Metrics	Performance metrics for services, Postgres databases and key-value stores: CPU and memory usage, limits and autoscaling targets, instance count, HTTP request count and latency quantiles, bandwidth and active connections	Read	Read-only. Queries are limited to the last 30 days with a minimum resolution of 30 seconds

Connect Render MCP to your AI agent

Four steps, under ten minutes.

01
Sign in to NeuralTrust
Open the NeuralTrust platform, sign in and go to Getting started, then Connect your tools.
02
Choose your agent
When asked which agent you use, pick Claude Code, Codex, Cursor, Windsurf, Copilot or your own code.
03
Connect Render
Find Render in the tool list and connect it.
04
Connect your agent
Follow the steps on screen to add the config to your agent, then send the test prompt.

Permissions down to the single tool

You do not allow Render MCP or block it. You allow the exact tools each agent can call, one by one.

Render MCP	26 TOOLS AVAILABLE
Code review agent	4 tools, read only
Issue triage bot	9 tools
Platform team assistant	all 26 tools

Tool by toolConnecting the server is not the unit of access. Each tool is granted on its own.
Default denyA tool you have not enabled is never exposed to the agent.
Approval on writesTools that change or delete things can require a human to confirm.

Is Render MCP secure?

One endpoint instead of raw credentials. Every call is checked against policy and logged.

AI agentsConnect through your MCP client configuration.	NeuralTrustApply controls based on your configuration.	Render MCP26 documented tools. Three of them named update_ only return a Dashboard link rather than changing anything.
Coding assistants	Connection & credentialsper user	Services9 tools
Internal copilots	Tool access policiesdeny by default	Postgres4 tools
Workflow automations	Request & response inspectioninline	Deployments4 tools
Compatible MCP clients	Activity monitoringalerts and logs	All other tools9 tools