Skip to content
actionlake
Start free

Connectors · Developer & infrastructure

Skyvern MCP server for AI agents.

actionlake is an MCP gateway: one remote MCP server that connects Claude and other MCP clients to Skyvern and the rest of your systems. It gives your agent 91 Skyvern actions (39 reads, 37 writes and 15 that wait for a confirmation) and holds the Skyvern credential itself, so the agent never does. It paces Skyvern's limits and records every call.

Connect Skyvern 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: "skyvern.agent_runs.list",  intent: "check which Skyvern account is connected"}# Get all runs · read · 1 upstream call
  • 91actionsin one signed file
  • 39readlook, never change
  • 37writecreate and update
  • 15wait for a yesdestructive, confirmed first

Set up

Connect Skyvern 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 Skyvern.

    API key. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead. actionlake stores what you paste write-only, sealed under your tenant's key.

    • api_keySkyvern API key: in the Skyvern app open Settings and copy the API key
    account { op: "connections.requirements", service: "skyvern" }
# 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: "skyvern.agent_runs.list" }# get all runs · read · 1 upstream callcheck { action: "skyvern.folders.create", input: { … } }# free: inputs, permissions, cost and confirmations, nothing sent

Skyvern at a glance

Sign-in
API key. No consent screen: your agent can connect it for you, but the key then passes through your chat. To keep it out of the model's context, paste it in the console instead.
You provide
api_key
Limits we pace to
60 calls a minute, per connection
API covered
Skyvern API 1.0.0 (OpenAPI 3.1.0)api.skyvern.com
Built from
Skyvern's documentation, read 23 October 2026api.skyvern.comdocs.skyvern.com
Service file
Version 1.0.0, signed (Ed25519)
Cached reads
29 of 39 reads, for up to 5 minutes; a write clears what it affects
Shared entities
1, so ask can line Skyvern up with your other systems
Data
Processed and stored in Sydney, Australia. No US region yet

Capabilities

What an AI agent can do in Skyvern.

91 actions in 21 areas. Every one carries its effect, so a key can allow the reads and hold the rest.

  • 39readlook, never change
  • 37writecreate and update
  • 15destructivewait for a confirmation

  1. Schedules

    9

    3 read · 4 write · 2 destructive

    Create a schedule for an agent · List all schedules for the organization · List schedules for an agent · Disable an agent schedule

  2. Agents

    8

    3 read · 3 write · 2 destructive

    Create a new agent · Get all agents · Get agent versions · Get an agent by id

  3. Run tags

    7

    5 read · 1 write · 1 destructive

    Batch get run tags · Batch get run tags (POST) · Get run tag history · Apply run tags

  4. Runs

    7

    4 read · 1 write · 2 destructive

    Get run timeline · List runs · Get artifacts for a run · Replay a run webhook

  5. Agent tags

    6

    4 read · 1 write · 1 destructive

    Batch get agent tags · Batch get agent tags (POST) · Get agent tag history · Apply agent tags

  6. Browser profiles

    6

    3 read · 2 write · 1 destructive

    Create a browser profile · List browser profiles · Get browser profile · Get browser profile usage

  7. Browser sessions

    6

    2 read · 3 write · 1 destructive

    Create a session · Get active browser sessions · Extend a session · Get a session

  8. Credentials

    5

    2 read · 2 write · 1 destructive

    Create credential · Get all credentials · Get credential by ID · Update credential

13 more areas · 37 actions
  • Folders 5
  • Scripts 5
  • Tag values 5
  • Agent runs 4
  • Custom LLMS 4
  • Job recipes 3
  • Tag keys 3
  • Tasks 3
  • Audit events 1
  • Browser types 1
  • SDK 1
  • Server 1
  • Totp codes 1

Top actions

The Skyvern API, shaped for agents.

The ids your agent types, what each does, and the one name it shares with other systems. 5 of Skyvern's 5 action names are shared with at least one other connector, so an agent learns them once.

ActionWhat it doesEffectSame name in
skyvern.folders.listGet foldersGet folders.readcontent.folder.list62 connectors
skyvern.folders.createCreate folderCreate folder.writecontent.folder.create62 connectors
skyvern.folders.deleteDelete folderDelete folder.destructivecontent.folder.delete46 connectors
skyvern.runs.listList runsList runs.readthis connector
skyvern.tasks.runRun a taskRun a task.writethis connector
skyvern.agents.listGet all agentsGet all agents.readthis connector
skyvern.agents.createCreate a new agentCreate a new agent.writethis connector
skyvern.agent_runs.runRun an agentRun an agent.writethis connector
skyvern.agent_runs.listGet all runsGet all runs.readthis connector
skyvern.tasks.run_loginLogin TaskLogin Task.writethis connector
skyvern.credentials.listGet all credentialsGet all credentials.readthis connector
skyvern.credentials.createCreate credentialCreate credential.writethis connector

Try it

Prompts to paste into Claude.

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

  1. Readread

    Get all runs in Skyvern and tell me what stands out.

    callsskyvern.agent_runs.list

  2. Writewrite

    Create folder in Skyvern. Show me exactly what you will send before you save it.

    callsskyvern.folders.create

Governance for Skyvern

Destructive actions wait for a yes.

15 Skyvern actions are destructive. These are all of them:

  • Delete folderskyvern.folders.delete
  • Delete an agentskyvern.agents.delete
  • Delete credentialskyvern.credentials.delete
  • Reset persisted browser profileskyvern.agents.reset_browser_profile
  • Cancel a run by idskyvern.runs.cancel
  • Bulk cancel runsskyvern.runs.bulk_cancel
  • Close a sessionskyvern.browser_sessions.close
  • Delete run tagskyvern.run_tags.delete
  • Delete tag keyskyvern.tag_keys.delete
  • Delete an agent scheduleskyvern.schedules.delete
  • Delete agent tagskyvern.agent_tags.delete
  • Delete tag valueskyvern.tag_values.delete
  • Delete Custom LLMskyvern.custom_llms.delete
  • Delete browser profileskyvern.browser_profiles.delete
  • Cancel a one-time agent scheduleskyvern.schedules.cancel

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.

doskyvern.folders.deleteconfirm_policy: human

Delete folder?

intent: the user asked me to delete folder

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

Cost

What Skyvern costs on actionlake.

Every Skyvern action is one upstream call, the only thing actionlake bills. Searching, check, cost, account and any call Skyvern 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 91 Skyvern actions
  • 91actions pageeach page of a long list is one call
  • 29reads 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 Skyvern action.

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

Schedules9
  • skyvern.schedules.cancelCancel a one-time agent scheduledestructive
  • skyvern.schedules.createCreate a schedule for an agentwrite
  • skyvern.schedules.deleteDelete an agent scheduledestructive
  • skyvern.schedules.disableDisable an agent schedulewrite
  • skyvern.schedules.enableEnable an agent schedulewrite
  • skyvern.schedules.getGet an agent schedule by idread
  • skyvern.schedules.listList all schedules for the organizationread
  • skyvern.schedules.list_for_agentList schedules for an agentread
  • skyvern.schedules.updateUpdate an agent schedulewrite
Agents8
  • skyvern.agents.createCreate a new agentwrite
  • skyvern.agents.deleteDelete an agentdestructive
  • skyvern.agents.getGet an agent by idread
  • skyvern.agents.listGet all agentsread
  • skyvern.agents.list_versionsGet agent versionsread
  • skyvern.agents.reset_browser_profileReset persisted browser profiledestructive
  • skyvern.agents.updateUpdate an agentwrite
  • skyvern.agents.update_folderUpdate agent folderwrite
Run tags7
  • skyvern.run_tags.applyApply run tagswrite
  • skyvern.run_tags.batch_getBatch get run tagsread
  • skyvern.run_tags.batch_get_postBatch get run tags (POST)read
  • skyvern.run_tags.deleteDelete run tagdestructive
  • skyvern.run_tags.getGet run tagsread
  • skyvern.run_tags.get_historyGet run tag historyread
  • skyvern.run_tags.get_suggestionsList run tag suggestionsread
Runs7
  • skyvern.runs.bulk_cancelBulk cancel runsdestructive
  • skyvern.runs.cancelCancel a run by iddestructive
  • skyvern.runs.getGet run info by idread
  • skyvern.runs.get_timelineGet run timelineread
  • skyvern.runs.listList runsread
  • skyvern.runs.list_artifactsGet artifacts for a runread
  • skyvern.runs.retry_webhookReplay a run webhookwrite
Agent tags6
  • skyvern.agent_tags.applyApply agent tagswrite
  • skyvern.agent_tags.batch_getBatch get agent tagsread
  • skyvern.agent_tags.batch_get_postBatch get agent tags (POST)read
  • skyvern.agent_tags.deleteDelete agent tagdestructive
  • skyvern.agent_tags.getGet agent tagsread
  • skyvern.agent_tags.get_historyGet agent tag historyread
Browser profiles6
  • skyvern.browser_profiles.createCreate a browser profilewrite
  • skyvern.browser_profiles.deleteDelete browser profiledestructive
  • skyvern.browser_profiles.getGet browser profileread
  • skyvern.browser_profiles.get_usageGet browser profile usageread
  • skyvern.browser_profiles.listList browser profilesread
  • skyvern.browser_profiles.updateUpdate browser profilewrite
Browser sessions6
  • skyvern.browser_sessions.closeClose a sessiondestructive
  • skyvern.browser_sessions.createCreate a sessionwrite
  • skyvern.browser_sessions.extendExtend a sessionwrite
  • skyvern.browser_sessions.getGet a sessionread
  • skyvern.browser_sessions.listGet active browser sessionsread
  • skyvern.browser_sessions.updateUpdate a sessionwrite
Credentials5
  • skyvern.credentials.createCreate credentialwrite
  • skyvern.credentials.deleteDelete credentialdestructive
  • skyvern.credentials.getGet credential by IDread
  • skyvern.credentials.listGet all credentialsread
  • skyvern.credentials.updateUpdate credentialwrite
Folders5
  • skyvern.folders.createCreate folderwrite
  • skyvern.folders.deleteDelete folderdestructive
  • skyvern.folders.getGet folderread
  • skyvern.folders.listGet foldersread
  • skyvern.folders.updateUpdate folderwrite
Scripts5
  • skyvern.scripts.createCreate scriptwrite
  • skyvern.scripts.deployDeploy scriptwrite
  • skyvern.scripts.getGet script by IDread
  • skyvern.scripts.listGet all scriptsread
  • skyvern.scripts.runRun scriptwrite
Tag values5
  • skyvern.tag_values.createCreate tag valuewrite
  • skyvern.tag_values.deleteDelete tag valuedestructive
  • skyvern.tag_values.listList tag valuesread
  • skyvern.tag_values.renameRename tag valuewrite
  • skyvern.tag_values.updateUpdate tag value colorwrite
Agent runs4
  • skyvern.agent_runs.listGet all runsread
  • skyvern.agent_runs.list_for_agentGet all runs by agentread
  • skyvern.agent_runs.retryRetry a runwrite
  • skyvern.agent_runs.runRun an agentwrite
Custom LLMS4
  • skyvern.custom_llms.createCreate Custom LLMwrite
  • skyvern.custom_llms.deleteDelete Custom LLMdestructive
  • skyvern.custom_llms.listList Custom LLMsread
  • skyvern.custom_llms.updateUpdate Custom LLMwrite
Job recipes3
  • skyvern.job_recipes.applyApply to a jobwrite
  • skyvern.job_recipes.extractExtract a job application postingwrite
  • skyvern.job_recipes.list_runsList job recipe runsread
Tag keys3
  • skyvern.tag_keys.deleteDelete tag keydestructive
  • skyvern.tag_keys.listList tag keysread
  • skyvern.tag_keys.updateUpdate tag keywrite
Tasks3
  • skyvern.tasks.runRun a taskwrite
  • skyvern.tasks.run_download_filesFile Download Taskwrite
  • skyvern.tasks.run_loginLogin Taskwrite
Audit events1
  • skyvern.audit_events.exportExport organization audit eventsread
Browser types1
  • skyvern.browser_types.listList selectable browser typesread
SDK1
  • skyvern.sdk.run_actionRun an SDK actionwrite
Server1
  • skyvern.server.get_versionGet server versionread
Totp codes1
  • skyvern.totp_codes.sendSend TOTP codewrite

Questions

The Skyvern MCP server, answered.

Is there a Skyvern MCP server for Claude?

Yes. actionlake is an MCP gateway: a remote MCP server that gives Claude, or another MCP client, 91 Skyvern actions across schedules, agents, run tags and runs, through the same eight tools it uses for every other connector. actionlake is independent; it is not made or endorsed by Skyvern.

What do I need to connect Skyvern?

Skyvern API key: in the Skyvern app open Settings and copy the API key. The free call account { op: "connections.requirements", service: "skyvern" } lists exactly what to paste.

Can an AI agent delete or change things in Skyvern?

Only if its key allows it, and only with an explicit confirmation. 15 Skyvern actions are destructive, including delete folder, delete an agent and delete credential. 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 Skyvern API limit?

actionlake paces every call to the limits in the Skyvern service file: 60 calls a minute per connection. Repeated reads are answered from cache (29 Skyvern actions can be), which never reaches Skyvern.

What does a Skyvern action cost?

One upstream call, the only unit actionlake bills. A list that pages bills one call per page (91 Skyvern actions page). Cache hits, searches, dry runs and calls Skyvern 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 Skyvern API does it cover?

The Skyvern API 1.0.0 (OpenAPI 3.1.0) at api.skyvern.com. It is built from Skyvern's documentation as read on 23 October 2026. 91 actions in 21 areas, each one in the signed service file (version 1.0.0).

Does actionlake keep my Skyvern 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 Skyvern 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 Skyvern

Shares the most action names with

Joins with, on shared entities

Browse

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

Give your agent Skyvern today.

Connect Skyvern freeSee pricing

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