Connectors · AI

Granola MCP server for AI agents.

actionlake is an MCP gateway: one remote MCP server that connects Claude and other MCP clients to Granola and the rest of your systems. It gives your agent 18 Granola actions (10 reads, 5 writes and 3 that wait for a confirmation) and holds the Granola credential itself, so the agent never does. It paces Granola's limits and records every call.

Connect Granola free See 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

18actionsin one signed file
10readlook, never change
5writecreate and update
3wait for a yesdestructive, confirmed first

Set up

Connect Granola in three steps.

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
02
Connect Granola.
Bearer token. 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_keyGranola API key (starts grn_)
```
account { op: "connections.requirements", service: "granola" }
# free: exactly what to paste, and what is found for you
```

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: "granola.folders.list" }# list folders · read · 1 upstream callcheck { action: "granola.webhook_endpoints.create", input: { … } }# free: inputs, permissions, cost and confirmations, nothing sent

Granola at a glance

Sign-in: Bearer token. 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: 5 calls a second, per connection
API covered: Granola public API v1 (OpenAPI 1.0.0)public-api.granola.ai
Built from: Granola's documentation, read 23 October 2026docs.granola.ai docs.granola.ai
Service file: Version 1.0.0, signed (Ed25519)
Cached reads: 10 of 10 reads, for up to 5 minutes; a write clears what it affects
Shared entities: 1, so ask can line Granola up with your other systems
Data: Processed and stored in Sydney, Australia. No US region yet

Capabilities

What an AI agent can do in Granola.

18 actions in 1 area. Every one carries its effect, so a key can allow the reads and hold the rest.

10readlook, never change
5writecreate and update
3destructivewait for a confirmation

Meeting notes
18
10 read · 5 write · 3 destructive
List folders · Create a webhook endpoint · List webhook endpoints · Update a webhook endpoint

Top actions

The Granola API, shaped for agents.

The ids your agent types, what each does, and the one name it shares with other systems. 7 of Granola's 7 action names are shared with at least one other connector, so an agent learns them once.

Action	What it does	Effect	Same name in
`granola.folders.list`List folders	List folders.	read	`content.folder.list`62 connectors
`granola.webhook_endpoints.create`Create a webhook endpoint	Create webhook endpoint.	write	`automation.webhook.create`209 connectors
`granola.webhook_endpoints.list`List webhook endpoints	List webhook endpoints.	read	`automation.webhook.list`202 connectors
`granola.webhook_endpoints.delete`Delete a webhook endpoint	Delete webhook endpoint.	destructive	`automation.webhook.delete`211 connectors
`granola.audit_events.list`List audit events	List audit events.	read	`identity.audit_event.list`51 connectors
`granola.transcripts.get`Get a note's transcript	Get transcript.	read	`comm.meeting_transcript.get`11 connectors
`granola.notes.list`List notes	List notes.	read	this connector
`granola.spaces.list`List spaces	List spaces.	read	this connector
`granola.legal_holds.list`List legal holds	List legal holds.	read	this connector
`granola.legal_holds.create`Create a legal hold	Create legal hold.	write	this connector
`granola.legal_hold_custodians.add`Add custodians to a legal hold	Add legal hold custodians.	write	this connector
`granola.legal_hold_custodians.list`List a legal hold's custodians	List legal hold custodians.	read	this connector

Try it

Prompts to paste into Claude.

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

Readread
List folders in Granola and tell me what stands out.
callsgranola.folders.list
Writewrite
Create a webhook endpoint in Granola. Show me exactly what you will send before you save it.
callsgranola.webhook_endpoints.create

Governance for Granola

Destructive actions wait for a yes.

3 Granola actions are destructive. These are all of them:

Delete a webhook endpointgranola.webhook_endpoints.delete
Release a legal holdgranola.legal_holds.release
Remove a custodian from a legal holdgranola.legal_hold_custodians.remove

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.

dogranola.webhook_endpoints.deleteconfirm_policy: human

Delete a webhook endpoint?

intent: the user asked me to delete a webhook endpoint

run suspended · waiting for a named person · key granola-agent cannot confirm itself

In the console, or from an emailed signed link.

a key for this connector

account {  op: "keys.create", name: "granola-agent",  allowed_actions: ["granola.*"], cap_calls: 2000,  confirm_policy: "human_for_writes",  approval_rules: [    { match: "granola.webhook_endpoints.delete", approver: "[email protected]" },    { match: "granola.legal_holds.release", approver: "[email protected]" },    { match: "granola.legal_hold_custodians.remove", approver: "[email protected]" }  ]}

Illustrative. Scoped to Granola only, capped at 2,000 upstream calls, every write confirmed, and the destructive ones approved by a named person.

Cost

What Granola costs on actionlake.

Every Granola action is one upstream call, the only thing actionlake bills. Searching, check, cost, account and any call Granola 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 18 Granola actions
18actions pageeach page of a long list is one call
10reads 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 Granola action.

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

Meeting notes18

granola.audit_events.listList audit eventsread
granola.folders.listList foldersread
granola.legal_hold_custodians.addAdd custodians to a legal holdwrite
granola.legal_hold_custodians.listList a legal hold's custodiansread
granola.legal_hold_custodians.removeRemove a custodian from a legal holddestructive
granola.legal_holds.createCreate a legal holdwrite
granola.legal_holds.getGet a legal holdread
granola.legal_holds.listList legal holdsread
granola.legal_holds.releaseRelease a legal holddestructive
granola.legal_holds.updateUpdate a legal holdwrite
granola.notes.getGet a noteread
granola.notes.listList notesread
granola.spaces.listList spacesread
granola.transcripts.getGet a note's transcriptread
granola.webhook_endpoints.createCreate a webhook endpointwrite
granola.webhook_endpoints.deleteDelete a webhook endpointdestructive
granola.webhook_endpoints.listList webhook endpointsread
granola.webhook_endpoints.updateUpdate a webhook endpointwrite

Questions

The Granola MCP server, answered.

Is there a Granola MCP server for Claude?

Yes. actionlake is an MCP gateway: a remote MCP server that gives Claude, or another MCP client, 18 Granola actions, through the same eight tools it uses for every other connector. actionlake is independent; it is not made or endorsed by Granola.

What do I need to connect Granola?

Granola API key (starts grn_). The free call account { op: "connections.requirements", service: "granola" } lists exactly what to paste.

Can an AI agent delete or change things in Granola?

Only if its key allows it, and only with an explicit confirmation. 3 Granola actions are destructive, including delete a webhook endpoint, release a legal hold and remove a custodian from a legal hold. 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 Granola API limit?

actionlake paces every call to the limits in the Granola service file: 5 calls a second per connection. Repeated reads are answered from cache (10 Granola actions can be), which never reaches Granola.

What does a Granola action cost?

One upstream call, the only unit actionlake bills. A list that pages bills one call per page (18 Granola actions page). Cache hits, searches, dry runs and calls Granola 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 Granola API does it cover?

The Granola public API v1 (OpenAPI 1.0.0) at public-api.granola.ai. It is built from Granola's documentation as read on 23 October 2026. 18 actions in 1 area, each one in the signed service file (version 1.0.0).

Does actionlake keep my Granola 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 Granola 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.

Give your agent Granola today.