Connectors
Granola

Granola

Live

OAUTH 2.1

MEETING NOTES

Granola MCP gives agents cited access to your meeting notes. Ask in plain language: every answer traces back to source.

  • Per-user credentials - each call uses the actual user’s token, never a shared bot.
  • Encrypted per-tenant vault  - AES-256, resolved at request time, never in LLM context.
  • Scope-checked & audited  -  pre-call scope check, 90-day SIEM-exportable audit chain.
Granola
agent · Acme Q3
Run
What did we commit to Acme last quarter?
S
granolamcp_query_granola_meetings
85ms
Granola agent
3 commits: 50-seat expansion, security review by Oct 30, custom SLO doc.
Sources: Oct 8, Oct 15, Oct 22
granolamcp
3 meetings
18:29
Draft a renewal brief for the CSM.
S
Message Claude...
Start in your coding agent
Up and running in one command
Install the Scalekit skill in your editor of choice. Connector, auth, tools, prompt, all wired up
Claude Code REPL
/plugin marketplace add scalekit-inc/claude-code-authstack
/plugin installagentkit@scalekit-auth-stack
Cursor Code REPL
/plugin marketplace add scalekit-inc/claude-code-authstack
/plugin installagentkit@scalekit-auth-stack
Windsurf Code REPL
/plugin marketplace add scalekit-inc/claude-code-authstack
/plugin installagentkit@scalekit-auth-stack
Codex Code REPL
/plugin marketplace add scalekit-inc/claude-code-authstack
/plugin installagentkit@scalekit-auth-stack
Copilot Code REPL
/plugin marketplace add scalekit-inc/claude-code-authstack
/plugin installagentkit@scalekit-auth-stack
CALL ANY GRANOLA TOOL
Every meeting action your agent needs, ready to call without writing auth or plumbing. Query, list, fetch, and transcribe, one toolkit, every framework
granolamcp_query_granola_meetings
Query meetings
Query Granola about the user's meetings using natural language. Returns a tailored response with inline citation links referencing source meeting notes. Citations must be preserved — they let users verify information by clicking through to original notes.
Parameters
Name
Type
Required
Description
query
string
Required
Natural language question (e.g. "What were my action items from last week?")
document_ids
array
Optional
Optional array of specific meeting IDs to scope the query
schema_version
string
Optional
Optional schema version.
tool_version
string
Optional
Optional tool version.
granolamcp_list_meetings
List meetings
granolamcp_get_meetings
Get meetings
granolamcp_get_meeting_transcript
Get transcript
Build your GRanola Agent
Three lines of code and your agent is talking to Granola. Drop in the toolkit, point it at a user, and every meeting becomes an action your agent can take
Python · LlamaIndex
1import os
2from scalekit import ScalekitClient
3from langchain.agents import create_tool_calling_agent
4
5client = ScalekitClient(
6    env_url=os.environ["SCALEKIT_ENV_URL"],
7    client_id=os.environ["SCALEKIT_CLIENT_ID"],
8    client_secret=os.environ["SCALEKIT_CLIENT_SECRET"],
9)
10
11# LangChain-compatible tools scoped to this user
12tools = client.actions.langchain.get_tools(
13    identifier="user@example.com",
14    connection_names=["granolamcp"],
15)
16
17agent = create_tool_calling_agent(llm, tools, prompt)
Python · LlamaIndex
1import os, openai
2from scalekit import ScalekitClient
3
4client = ScalekitClient(
5    env_url=os.environ["SCALEKIT_ENV_URL"],
6    client_id=os.environ["SCALEKIT_CLIENT_ID"],
7    client_secret=os.environ["SCALEKIT_CLIENT_SECRET"],
8)
9
10# Tool schemas in OpenAI format
11schemas = client.actions.tools.list_scoped_tools(
12    identifier="user@example.com",
13    page_size=100,
14).tools
15
16response = openai.chat.completions.create(
17    model="gpt-4o",
18    tools=schemas,
19    messages=[{"role": "user", "content": "What did we commit to Acme last quarter?"}],
20)
21
22# On tool_call, route back through Scalekit:
23# client.actions.execute_tool(
24#     tool_name=tc.function.name,
25#     tool_input=tc.function.arguments,
26#     identifier="user@example.com",
27# )
Python · LlamaIndex
1import os, anthropic
2from scalekit import ScalekitClient
3
4client = ScalekitClient(
5    env_url=os.environ["SCALEKIT_ENV_URL"],
6    client_id=os.environ["SCALEKIT_CLIENT_ID"],
7    client_secret=os.environ["SCALEKIT_CLIENT_SECRET"],
8)
9
10schemas = client.actions.tools.list_scoped_tools(
11    identifier="user@example.com",
12    page_size=100,
13).tools
14
15ac = anthropic.Anthropic()
16response = ac.messages.create(
17    model="claude-opus-4-7",
18    tools=schemas,
19    messages=[{"role": "user", "content": "Summarise my meetings with Acme last month"}],
20)
21
22# On tool_use block, route back through Scalekit:
23# client.actions.execute_tool(
24#     tool_name=block.name,
25#     tool_input=block.input,
26#     identifier="user@example.com",
27# )
Python · LlamaIndex
1import os
2from scalekit import ScalekitClient
3from google.adk.agents import LlmAgent
4
5client = ScalekitClient(
6    env_url=os.environ["SCALEKIT_ENV_URL"],
7    client_id=os.environ["SCALEKIT_CLIENT_ID"],
8    client_secret=os.environ["SCALEKIT_CLIENT_SECRET"],
9)
10
11tools = client.actions.langchain.get_tools(
12    identifier="user@example.com",
13    connection_names=["granolamcp"],
14)
15
16agent = LlmAgent(
17    model="gemini-2.0-flash",
18    tools=tools,
19    instruction="You are a meeting intelligence assistant.",
20)
Try these prompts
Copy any prompt into your agent. Each maps directly to a Granola tool — click to copy, paste into your agent, done.
Action & follow-up
Copy the prompt
Copied
What action items came out of my meetings this week?
Copy the prompt
Copied
What did we commit to Acme last quarter?
Copy the prompt
Copied
What are the open questions from yesterday's meeting?
Copy the prompt
Copied
Draft a follow-up email based on my last meeting with Stripe.
Search & recall
Copy the prompt
Copied
Find every meeting where pricing was discussed.
Copy the prompt
Copied
What did [person] say about the roadmap?
Copy the prompt
Copied
Summarise everything discussed about [feature] across all meetings.
Copy the prompt
Copied
List all my meetings from last month.
Prep & intel
Copy the prompt
Copied
What context do I need before my next call with this account?
Copy the prompt
Copied
Who attended my sync with [company] and what was decided?
Copy the prompt
Copied
Get the full transcript from my most recent board meeting.
SEE HOW AUTH WORKS
Your users connect once. Their Granola credentials stay vaulted, every call is checked, and every action is logged.
1
Authorize
Your user connects
Granola
once. We tie it to their identity and the meetings they approved — no shared bot account, no org-wide access
Who:
user ‘A’
when:
Once per user
access:
A’s meetings only
2
Store
Their
Granola
token lives in a vault scoped to them. User A's meetings are never reachable by an agent acting for user B, even on the same connection
vault:
encrypted
scope:
per-user
tokens:
auto-refreshed
3
Resolve
When your agent calls a
Granola
tool, we fetch the right token server-side. It never touches your agent, never appears in the LLM context, never shows up in your logs
speed:
~40ms
check:
before every call
seen by:
nobody
4
Audit
Every
Granola
tool call is logged — who triggered it, which meeting was fetched, what came back. 90 days of history, tied to the user who authorized it
history:
90 days
export:
SIEM-ready
logged:
every call
Test other agents
Real agents solving real workflows, each with working code and a live demo you can run in production today. Pick a use case and build from something already shipping
SALES
Deal Intelligence Agent
Combine Gong call recordings, Attio contact data, and Slack signals to surface deal risks and next-best actions. Updated after every call.
GTM
Meeting intelligence report
Sync Granola meeting summaries into Salesforce records and push status updates to Slack. No manual CRM entry.
Sales
Real actions, on behalf of real users
Pull meeting notes from Granola, cross-reference contact history in Attio, and generate a pre-call brief before every sales meeting. Zero rep input.
DEAL OPS
Renewal risk briefing
Build targeted prospect lists with Apollo, enrich with Granola conversation history, and draft personalised outreach. Runs on a schedule.
Why Scalekit
Secure your agent's access. Connectors ship in minutes
A lot of agent connectivity tools were built for demos or retrofitted APIs. Scalekit starts with identity, scope enforcement, and audit — and connectors follow.
01.
Acting as a bot fails in production
A shared Granola token looks fine in a demo. In production, every note retrieved looks like it came from a service account. Audit logs break. Per-user scoping breaks. Scalekit resolves the credential of the actual user who triggered the agent — never a shared bot.
// shared bot token
token = "sk_shared_xxx"
audit → bot
user_filter → broken

// scalekit · per-user
token = resolve(user_id)
audit → user_abc
scope → enforced ✓
02.
Authentication is not authorization
03.
Multi-tenancy is architectural
04.
The N×M connector problem
Capability
DIY
Scalekit AgentKit
Token storage
Build + maintain yourself
AES-256 vault, managed
Per-user isolation
Custom credential map
Per-tenant namespace, default
Scope enforcement
Manual checks or none
Per-request, pre-API call
Token refresh
Cron job you maintain
Automatic
Audit trail
Build your own logging
90-day, SIEM-exportable
New connector
New OAuth implementation
Same pattern, one config
Multi-framework
Per-framework adapter code
8 adapters included
“Our agents act across Salesforce, Gong, Google Drive, and more, on behalf of every customer. Scalekit behind the scenes meant we can keep adding tools without ever rebuilding how credentials or tool calling work.”
Venu Madhav Kattagoni
Head of Engineering / Von