Skip to content
actionlake
Start free

Connectors · Developer & infrastructure

Google Cloud Run MCP server for AI agents.

actionlake is an MCP gateway: one remote MCP server that connects Claude and other MCP clients to Google Cloud Run and the rest of your systems. It gives your agent 58 Google Cloud Run actions (32 reads, 12 writes and 14 that wait for a confirmation) and holds the Google Cloud Run credential itself, so the agent never does. It paces Google Cloud Run's limits and records every call. Connecting Google Cloud Run needs your own Google Cloud Run OAuth app and one admin sign-in.

Connect Google Cloud Run freeSee pricing

100,000 free calls a month. No card, no email, no password: press Start a free sandbox on the next screen and keep the key it shows you.

claude mcp add --transport http actionlake https://connect.actionlake.com/mcp
the first call
do {  action: "gcp_cloud_run.locations.export_project_metadata",  intent: "check which Google Cloud Run account is connected"}# Projects locations: export project metadata · read · 1 upstream call
  • 58actionsin one signed file
  • 32readlook, never change
  • 12writecreate and update
  • 14wait for a yesdestructive, confirmed first

Set up

Connect Google Cloud Run in three steps.

  1. 01

    Add actionlake to Claude.

    Run the command once in Claude Code; its consent page can create a free sandbox: 100,000 calls a month, no card. In the Claude app, add a custom connector with the URL below and paste your key on the consent page. Creating a new sandbox from the Claude app is not yet tested by us.

    https://connect.actionlake.com/mcp
  2. 02

    Connect Google Cloud Run.

    OAuth 2.0, authorization code with PKCE. You register your own OAuth app with Google Cloud Run, an admin signs in once, and you paste the app's client id and secret and the tokens from that sign-in; actionlake renews the access token from then on. actionlake stores what you paste write-only, sealed under your tenant's key.

    Before the first sign-in, Google Cloud Run needs an OAuth app of your own, registered in its developer portal. actionlake has no Google Cloud Run app to lend you.

    • access_tokenAccess token from one sign-in with your own OAuth app
    • refresh_tokenRefresh token from the same sign-in; the access token is renewed from it
    • client_idGoogle OAuth client id (Google Cloud console > APIs & Services > Credentials; enable the Cloud Run Admin API on the…
    • client_secretGoogle OAuth client secret
    account { op: "connections.requirements", service: "gcp_cloud_run" }
# free: exactly what to paste, and what is found for you
  3. 03

    Make the first call. Check before a write.

    Ask in plain words; Claude picks the action. Before anything changes, check validates the call and prices it, for free.

    do    { action: "gcp_cloud_run.locations.export_project_metadata" }# projects locations: export project metadata · read · 1 upstream callcheck { action: "gcp_cloud_run.jobs.create", input: { … } }# free: inputs, permissions, cost and confirmations, nothing sent

Google Cloud Run at a glance

Sign-in
OAuth 2.0, authorization code with PKCE. You register your own OAuth app with Google Cloud Run, an admin signs in once, and you paste the app's client id and secret and the tokens from that sign-in; actionlake renews the access token from then on.
You provide
access_tokenrefresh_tokenclient_idclient_secret
Limits we pace to
600 calls a minute, per connection
API covered
Google Cloud Run Admin API v2 (services, revisions, jobs, executions, tasks, worker pools, instances, operations) API run v2 (Discovery revision 20261016)run.googleapis.com
Built from
Google Cloud Run's documentation, read 23 October 2026run.googleapis.comcloud.google.com
Service file
Version 1.0.0, signed (Ed25519)
Cached reads
25 of 32 reads, for up to 5 minutes; a write clears what it affects
Shared entities
2, so ask can line Google Cloud Run up with your other systems
Data
Processed and stored in Sydney, Australia. No US region yet

Capabilities

What an AI agent can do in Google Cloud Run.

58 actions in 11 areas. Every one carries its effect, so a key can allow the reads and hold the rest.

  • 32readlook, never change
  • 12writecreate and update
  • 14destructivewait for a confirmation

  1. Instances

    10

    4 read · 3 write · 3 destructive

    Projects locations instances: create · Projects locations instances: list · Projects locations instances: test iam permissions · Projects locations instances: get

  2. Jobs

    9

    4 read · 3 write · 2 destructive

    Projects locations jobs: list · Projects locations jobs: create · Projects locations jobs: get · Projects locations jobs: patch

  3. Services

    8

    4 read · 2 write · 2 destructive

    Projects locations services: list · Projects locations services: create · Projects locations services: get · Projects locations services: patch

  4. Worker pools

    8

    4 read · 2 write · 2 destructive

    Projects locations worker pools: create · Projects locations worker pools: list · Projects locations worker pools: test iam permissions · Projects locations worker pools: get

  5. Jobs executions

    5

    3 read · 0 write · 2 destructive

    Projects locations jobs executions: list · Projects locations jobs executions: export status · Projects locations jobs executions: get · Projects locations jobs executions: delete

  6. Locations

    4

    3 read · 1 write · 0 destructive

    Projects locations: export image · Projects locations: export image metadata · Projects locations: export metadata · Projects locations: export project metadata

  7. Operations

    4

    3 read · 0 write · 1 destructive

    Projects locations operations: list · Projects locations operations: wait · Projects locations operations: get · Projects locations operations: delete

  8. Services revisions

    4

    3 read · 0 write · 1 destructive

    Projects locations services revisions: list · Projects locations services revisions: export status · Projects locations services revisions: get · Projects locations services revisions: delete

3 more areas · 6 actions
  • Worker pools revisions 3
  • Jobs executions tasks 2
  • Builds 1

Top actions

The Google Cloud Run API, shaped for agents.

The ids your agent types, what each does, and the one name it shares with other systems. 10 of Google Cloud Run's 10 action names are shared with at least one other connector, so an agent learns them once.

ActionWhat it doesEffectSame name in
gcp_cloud_run.jobs.listProjects locations jobs: listLists Jobs.readinfra.job.list12 connectors
gcp_cloud_run.services.listProjects locations services: listLists Services.readinfra.container_service.list11 connectors
gcp_cloud_run.jobs.createProjects locations jobs: createCreates a Job.writeinfra.job.create10 connectors
gcp_cloud_run.services.createProjects locations services: createCreates a new Service in a given project and location.writeinfra.container_service.create8 connectors
gcp_cloud_run.services.deleteProjects locations services: deleteDeletes a Service.destructiveinfra.container_service.delete11 connectors
gcp_cloud_run.jobs.runProjects locations jobs: runTriggers creation of a new Execution of this Job.writethis connector
gcp_cloud_run.operations.listProjects locations operations: listLists operations that match the specified filter in the request.readthis connector
gcp_cloud_run.operations.waitProjects locations operations: waitWaits until the specified long-running operation is done or reaches at most a specified timeout, returning the latest state.readthis connector
gcp_cloud_run.jobs.test_iam_permissionsProjects locations jobs: test iam permissionsReturns permissions that a caller has on the specified Project.readthis connector
gcp_cloud_run.services.test_iam_permissionsProjects locations services: test iam permissionsReturns permissions that a caller has on the specified Project.readthis connector
gcp_cloud_run.services_revisions.listProjects locations services revisions: listLists Revisions from a given Service, or from a given location.readthis connector
gcp_cloud_run.builds.submitProjects locations builds: submitSubmits a build in a given project.writethis connector

Try it

Prompts to paste into Claude.

Each one names the Google Cloud Run actions it will call, so you know what it reads and what it changes before you send it.

  1. Readread

    Projects locations: export project metadata in Google Cloud Run and tell me what stands out.

    callsgcp_cloud_run.locations.export_project_metadata

  2. Writewrite

    Projects locations jobs: create in Google Cloud Run. Show me exactly what you will send before you save it.

    callsgcp_cloud_run.jobs.create

Governance for Google Cloud Run

Destructive actions wait for a yes.

14 Google Cloud Run actions are destructive. These are all of them:

  • Projects locations services: deletegcp_cloud_run.services.delete
  • Projects locations jobs: deletegcp_cloud_run.jobs.delete
  • Projects locations operations: deletegcp_cloud_run.operations.delete
  • Projects locations jobs: set iam policygcp_cloud_run.jobs.set_iam_policy
  • Projects locations services: set iam policygcp_cloud_run.services.set_iam_policy
  • Projects locations jobs executions: deletegcp_cloud_run.jobs_executions.delete
  • Projects locations services revisions: deletegcp_cloud_run.services_revisions.delete
  • Projects locations jobs executions: cancelgcp_cloud_run.jobs_executions.cancel
  • Projects locations instances: stopgcp_cloud_run.instances.stop
  • Projects locations instances: deletegcp_cloud_run.instances.delete
  • Projects locations worker pools: deletegcp_cloud_run.worker_pools.delete
  • Projects locations instances: set iam policygcp_cloud_run.instances.set_iam_policy
  • Projects locations worker pools: set iam policygcp_cloud_run.worker_pools.set_iam_policy
  • Projects locations worker pools revisions: deletegcp_cloud_run.worker_pools_revisions.delete

A destructive action runs only with confirm: true. Under the human policy the agent cannot confirm for itself: the run stops until a named person approves, in the console or from a signed link in an email. The cap on a key is refused before the call is sent, never after.

dogcp_cloud_run.services.deleteconfirm_policy: human

Projects locations services: delete?

intent: the user asked me to projects locations services: delete

run suspended · waiting for a named person · key gcp-cloud-run-agent cannot confirm itself
In the console, or from an emailed signed link.
a key for this connector
account {  op: "keys.create", name: "gcp-cloud-run-agent",  allowed_actions: ["gcp_cloud_run.*"], cap_calls: 2000,  confirm_policy: "human_for_writes",  approval_rules: [    { match: "gcp_cloud_run.services.delete", approver: "[email protected]" },    { match: "gcp_cloud_run.jobs.delete", approver: "[email protected]" },    { match: "gcp_cloud_run.operations.delete", approver: "[email protected]" }  ]}
Illustrative. Scoped to Google Cloud Run only, capped at 2,000 upstream calls, every write confirmed, and the destructive ones approved by a named person.

Cost

What Google Cloud Run costs on actionlake.

Every Google Cloud Run action is one upstream call, the only thing actionlake bills. Searching, check, cost, account and any call Google Cloud Run refused with 401, 403, 429 or a 5xx are free. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000.

  • 1upstream call per actionall 58 Google Cloud Run actions
  • 9actions pageeach page of a long list is one call
  • 25reads can be cacheda cache hit costs nothing
  • 100,000calls a month, freepaid plans from US$49 a month for 250,000

See pricing

Action index

Every Google Cloud Run action.

58 actions by area: the id your agent calls, what it does, and its effect.

Instances10
  • gcp_cloud_run.instances.createProjects locations instances: createwrite
  • gcp_cloud_run.instances.deleteProjects locations instances: deletedestructive
  • gcp_cloud_run.instances.getProjects locations instances: getread
  • gcp_cloud_run.instances.get_iam_policyProjects locations instances: get iam policyread
  • gcp_cloud_run.instances.listProjects locations instances: listread
  • gcp_cloud_run.instances.patchProjects locations instances: patchwrite
  • gcp_cloud_run.instances.set_iam_policyProjects locations instances: set iam policydestructive
  • gcp_cloud_run.instances.startProjects locations instances: startwrite
  • gcp_cloud_run.instances.stopProjects locations instances: stopdestructive
  • gcp_cloud_run.instances.test_iam_permissionsProjects locations instances: test iam permissionsread
Jobs9
  • gcp_cloud_run.jobs.createProjects locations jobs: createwrite
  • gcp_cloud_run.jobs.deleteProjects locations jobs: deletedestructive
  • gcp_cloud_run.jobs.getProjects locations jobs: getread
  • gcp_cloud_run.jobs.get_iam_policyProjects locations jobs: get iam policyread
  • gcp_cloud_run.jobs.listProjects locations jobs: listread
  • gcp_cloud_run.jobs.patchProjects locations jobs: patchwrite
  • gcp_cloud_run.jobs.runProjects locations jobs: runwrite
  • gcp_cloud_run.jobs.set_iam_policyProjects locations jobs: set iam policydestructive
  • gcp_cloud_run.jobs.test_iam_permissionsProjects locations jobs: test iam permissionsread
Services8
  • gcp_cloud_run.services.createProjects locations services: createwrite
  • gcp_cloud_run.services.deleteProjects locations services: deletedestructive
  • gcp_cloud_run.services.getProjects locations services: getread
  • gcp_cloud_run.services.get_iam_policyProjects locations services: get iam policyread
  • gcp_cloud_run.services.listProjects locations services: listread
  • gcp_cloud_run.services.patchProjects locations services: patchwrite
  • gcp_cloud_run.services.set_iam_policyProjects locations services: set iam policydestructive
  • gcp_cloud_run.services.test_iam_permissionsProjects locations services: test iam permissionsread
Worker pools8
  • gcp_cloud_run.worker_pools.createProjects locations worker pools: createwrite
  • gcp_cloud_run.worker_pools.deleteProjects locations worker pools: deletedestructive
  • gcp_cloud_run.worker_pools.getProjects locations worker pools: getread
  • gcp_cloud_run.worker_pools.get_iam_policyProjects locations worker pools: get iam policyread
  • gcp_cloud_run.worker_pools.listProjects locations worker pools: listread
  • gcp_cloud_run.worker_pools.patchProjects locations worker pools: patchwrite
  • gcp_cloud_run.worker_pools.set_iam_policyProjects locations worker pools: set iam policydestructive
  • gcp_cloud_run.worker_pools.test_iam_permissionsProjects locations worker pools: test iam permissionsread
Jobs executions5
  • gcp_cloud_run.jobs_executions.cancelProjects locations jobs executions: canceldestructive
  • gcp_cloud_run.jobs_executions.deleteProjects locations jobs executions: deletedestructive
  • gcp_cloud_run.jobs_executions.export_statusProjects locations jobs executions: export statusread
  • gcp_cloud_run.jobs_executions.getProjects locations jobs executions: getread
  • gcp_cloud_run.jobs_executions.listProjects locations jobs executions: listread
Locations4
  • gcp_cloud_run.locations.export_imageProjects locations: export imagewrite
  • gcp_cloud_run.locations.export_image_metadataProjects locations: export image metadataread
  • gcp_cloud_run.locations.export_metadataProjects locations: export metadataread
  • gcp_cloud_run.locations.export_project_metadataProjects locations: export project metadataread
Operations4
  • gcp_cloud_run.operations.deleteProjects locations operations: deletedestructive
  • gcp_cloud_run.operations.getProjects locations operations: getread
  • gcp_cloud_run.operations.listProjects locations operations: listread
  • gcp_cloud_run.operations.waitProjects locations operations: waitread
Services revisions4
  • gcp_cloud_run.services_revisions.deleteProjects locations services revisions: deletedestructive
  • gcp_cloud_run.services_revisions.export_statusProjects locations services revisions: export statusread
  • gcp_cloud_run.services_revisions.getProjects locations services revisions: getread
  • gcp_cloud_run.services_revisions.listProjects locations services revisions: listread
Worker pools revisions3
  • gcp_cloud_run.worker_pools_revisions.deleteProjects locations worker pools revisions: deletedestructive
  • gcp_cloud_run.worker_pools_revisions.getProjects locations worker pools revisions: getread
  • gcp_cloud_run.worker_pools_revisions.listProjects locations worker pools revisions: listread
Jobs executions tasks2
  • gcp_cloud_run.jobs_executions_tasks.getProjects locations jobs executions tasks: getread
  • gcp_cloud_run.jobs_executions_tasks.listProjects locations jobs executions tasks: listread
Builds1
  • gcp_cloud_run.builds.submitProjects locations builds: submitwrite

Questions

The Google Cloud Run MCP server, answered.

Is there a Google Cloud Run MCP server for Claude?

Yes. actionlake is an MCP gateway: a remote MCP server that gives Claude, or another MCP client, 58 Google Cloud Run actions across instances, jobs, services and worker pools, through the same eight tools it uses for every other connector. actionlake is independent; it is not made or endorsed by Google Cloud Run.

What do I need to connect Google Cloud Run?

Your own Google Cloud Run OAuth app, registered with Google Cloud Run (its client id and client secret), and the access and refresh tokens from one admin sign-in. actionlake has no Google Cloud Run app of its own, so this is not a one-step consent. actionlake renews the access token as it expires. The free call account { op: "connections.requirements", service: "gcp_cloud_run" } lists exactly what to paste.

Can an AI agent delete or change things in Google Cloud Run?

Only if its key allows it, and only with an explicit confirmation. 14 Google Cloud Run actions are destructive, including projects locations services: delete, projects locations jobs: delete and projects locations operations: delete. Each runs only with confirm: true. Under the human confirm policy, or an approval rule on the key, the run waits until a named person approves it.

Will an AI agent use up my Google Cloud Run API limit?

actionlake paces every call to the limits in the Google Cloud Run service file: 600 calls a minute per connection. Repeated reads are answered from cache (25 Google Cloud Run actions can be), which never reaches Google Cloud Run.

What does a Google Cloud Run action cost?

One upstream call, the only unit actionlake bills. A list that pages bills one call per page (9 Google Cloud Run actions page). Cache hits, searches, dry runs and calls Google Cloud Run refused with 401, 403, 429 or a 5xx are free. The Free plan includes 100,000 calls a month with no card. Paid plans start at US$49 a month for 250,000 calls, with overage from US$0.30 down to US$0.18 per 1,000.

Which Google Cloud Run API does it cover?

The Google Cloud Run Admin API v2 (services, revisions, jobs, executions, tasks, worker pools, instances, operations) API run v2 (Discovery revision 20261016) at run.googleapis.com. It is built from Google Cloud Run's documentation as read on 23 October 2026. 58 actions in 11 areas, each one in the signed service file (version 1.0.0).

Does actionlake keep my Google Cloud Run data, and where?

actionlake processes and stores data in Sydney, Australia; there is no US region yet. Reads may be cached for up to 5 minutes, and a write clears the reads it affects. Each run's request and response is kept, sealed, for your plan's trace window: 1 day on Free, 3 days on Starter, 7 days on Pro, 14 days on Team, 30 days on Business and 30 days on Enterprise. Your Google Cloud Run credential is sealed under your tenant's key, never logged and never returned. Your traffic is not used to train models.

Which AI clients can use it?

Claude Code, tested by us end to end, and the Claude app as a custom connector, tested by us with an existing key. Other MCP clients that support remote servers over Streamable HTTP (MCP revisions 2025-06-18 and 2025-11-25) and OAuth dynamic client registration, or a bearer header, should connect the same way; we have not tested them. Or plain REST, under the same action names.

More connectors like Google Cloud Run

Shares the most action names with

Joins with, on shared entities

Browse

actionlake is independent. It is not made or endorsed by Google Cloud Run; the name is the vendor's own. Counts are from the Google Cloud Run service file, version 1.0.0.

Give your agent Google Cloud Run today.

Connect Google Cloud Run freeSee pricing

Start free - 100,000 calls/month, no card. No signup. Hard caps on every key. Cached answers are free.