Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@mcp-granola search for notes about quarterly planning"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
An MCP (Model Context Protocol) server that gives AI agents access to your Granola meeting notes. Search notes, retrieve full content with transcripts, and paginate through your meeting history.
List notes with date filtering and cursor-based pagination
Get full note content including AI summaries and meeting transcripts
Search notes by title or summary keywords
Works with Claude Code, Gemini CLI, and any MCP-compatible client
Zero external dependencies beyond the MCP SDK
Related MCP server: Bruin
npm install @devli13/mcp-granolaOr clone and install locally:
git clone https://github.com/devli13/mcp-granola.git
cd mcp-granola
npm installOpen the Granola desktop app
Go to Settings > API
Click Create new key
Copy the key (starts with grn_)
Add to your .mcp.json (Claude Code) or .gemini/settings.json (Gemini CLI):
{
"mcpServers": {
"granola": {
"command": "npx",
"args": ["-y", "@devli13/mcp-granola"],
"env": {
"GRANOLA_API_KEY": "grn_your_api_key_here"
}
}
}
}Or run directly:
GRANOLA_API_KEY=grn_your_key node server.jsEnvironment Variable | Required | Description |
| Yes | Your Granola API key (starts with |
Tool | Description | Key Parameters |
| List notes, most-recent first |
|
| Fetch a single note by ID |
|
| Search notes by title/summary keywords |
|
{
"name": "list_notes",
"arguments": {
"limit": 5,
"created_after": "2025-01-01T00:00:00Z"
}
}{
"name": "get_note",
"arguments": {
"id": "not_abc123def456",
"include_transcript": true
}
}Granola API is in beta — endpoints and response shapes may change.
Personal API keys only access notes you own or that are shared with you directly.
search_notes performs client-side substring matching over the 200 most recent notes. It is not a full-text search engine. For large note volumes, use list_notes with date filters instead.
The API only returns notes with generated AI summaries and transcripts. Notes still being processed won't appear.
Granola enforces rate limits per workspace:
Metric | Value |
Burst capacity | 25 requests |
Time window | 5 seconds |
Sustained rate | 5 req/s (300/min) |
Issues and pull requests are welcome! Please open an issue first to discuss any significant changes.
MIT
Fetch a Granola note by ID. Pass include="transcript" to also return the transcript.
| Name | Required | Description | Default |
|---|---|---|---|
| id | Yes | Note ID, like not_XXXXXXXXXXXXXX. | |
| include_transcript | No | Include the meeting transcript. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries the full burden of behavioral disclosure. It mentions fetching a note and including a transcript, but does not disclose other behavioral traits such as error handling (e.g., what happens if the ID is invalid), authentication needs, rate limits, or whether it's a read-only operation. This leaves significant gaps for an agent to understand how to use it safely.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is two sentences with zero waste, front-loaded with the core purpose and followed by a specific usage tip. Every sentence earns its place by adding actionable information, making it efficient and easy to parse.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given the tool's low complexity (2 parameters, no output schema, no annotations), the description is adequate but incomplete. It covers the basic purpose and a parameter tip, but lacks details on behavioral aspects like error handling or return values, which are important for a tool with no annotations or output schema. It's minimally viable but has clear gaps.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
The schema description coverage is 100%, so the schema already documents both parameters (id and include_transcript) with descriptions. The description adds minimal value by mentioning the include parameter's purpose ('to also return the transcript'), but does not provide additional syntax or format details beyond what the schema provides. This meets the baseline for high schema coverage.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the specific action ('Fetch') and resource ('a Granola note by ID'), distinguishing it from sibling tools like 'list_notes' (which lists multiple notes) and 'search_notes' (which searches notes). It precisely defines what the tool does without ambiguity.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear context on when to use the tool (to fetch a specific note by ID) and includes an optional feature (include transcript), but it does not explicitly state when not to use it or name alternatives like 'list_notes' or 'search_notes'. This gives good guidance but lacks explicit exclusions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
List Granola notes, most-recent first. Supports created_after/created_before ISO timestamps and a cursor for pagination.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | Max notes per page (default 20, max 100). | |
| created_after | No | ISO-8601 timestamp; only notes created strictly after this. | |
| created_before | No | ISO-8601 timestamp; only notes created strictly before this. | |
| cursor | No | Pagination cursor from previous response.hasMore. |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries the full burden. It discloses ordering behavior ('most-recent first'), filtering capabilities, and pagination support, but lacks details on rate limits, authentication needs, error handling, or response format, which are important for a list operation.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is a single, well-structured sentence that efficiently conveys purpose, ordering, and key features (filtering and pagination) without any wasted words, making it easy to parse and understand quickly.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
Given no annotations and no output schema, the description covers basic functionality but lacks details on response structure (e.g., what fields are returned), error cases, or advanced usage scenarios, leaving gaps for an AI agent to fully understand tool behavior.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema already documents all parameters thoroughly. The description adds context by mentioning ISO timestamps and cursor usage, but does not provide additional meaning beyond what the schema specifies, such as default values or constraints not in the schema.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the action ('List'), resource ('Granola notes'), and ordering ('most-recent first'), which is specific and distinguishes it from sibling tools like get_note (single note) and search_notes (likely keyword-based).
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description implies usage for retrieving multiple notes with time-based filtering and pagination, but does not explicitly state when to use this tool versus alternatives like search_notes, nor does it mention any prerequisites or exclusions.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Client-side substring search over the list_notes response (title + summary). Pages up to 200 most recent notes then filters.
| Name | Required | Description | Default |
|---|---|---|---|
| query | Yes | Case-insensitive substring to match on title/summary. | |
| limit | No | Max matches returned (default 10). |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations provided, the description carries full burden and discloses important behavioral traits: it's a client-side filter (not server-side search), operates on cached list_notes response, has a 200-note limit, returns up to limit matches, and performs case-insensitive substring matching. However, it doesn't mention performance characteristics or error conditions.
Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.
Is the description appropriately sized, front-loaded, and free of redundancy?
The description is perfectly concise - one sentence that packs essential information about scope, method, and limitations. Every word earns its place with zero waste, and it's front-loaded with the core functionality.
Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.
Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?
For a search tool with no annotations and no output schema, the description does well by explaining the client-side nature, 200-note limit, and substring matching. However, it doesn't describe the return format (what fields are included) or what happens when no matches are found, which would be helpful given the lack of output schema.
Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.
Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?
Schema description coverage is 100%, so the schema already fully documents both parameters. The description adds marginal value by mentioning the search operates on 'title + summary' and the 200-note context, but doesn't provide additional syntax or format details beyond what the schema provides.
Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.
Does the description clearly state what the tool does and how it differs from similar tools?
The description clearly states the specific action ('client-side substring search') and resource ('list_notes response') with precise scope ('title + summary'). It explicitly distinguishes from sibling tools by mentioning it operates on list_notes response rather than direct database access, unlike get_note or list_notes.
Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.
Does the description explain when to use this tool, when not to, or what alternatives exist?
The description provides clear context about when to use this tool ('client-side substring search over the list_notes response') and mentions the 200-note limit, but doesn't explicitly state when NOT to use it or name alternatives. It implies usage for substring matching rather than exact or structured queries.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Each tool has a clearly distinct purpose: get_note retrieves a single note by ID, list_notes lists notes with filtering and pagination, and search_notes performs substring searches over note content. There is no overlap in functionality, making tool selection straightforward for an agent.
All tool names follow a consistent verb_noun pattern (get_note, list_notes, search_notes) with clear, descriptive verbs. The naming is uniform and predictable, enhancing usability and reducing confusion.
With only 3 tools, the server feels thin for a note management domain, as it lacks essential CRUD operations like create, update, or delete. While the tools are well-defined, the count is borderline low for comprehensive functionality.
The tool surface is significantly incomplete for note management, covering only read and search operations. There are no tools for creating, updating, or deleting notes, which are critical for a full lifecycle, leading to potential agent failures in workflows requiring modification.
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Google Keep-style notes app with an MCP server for AI agents to read/write notes.
Markdown-based note-taking with a hosted MCP server. Your notes serve you and your AI.
Agent-native MCP server over the public saagarpatel.dev corpus. Read-only, stateless.
An MCP server that provides tools to discover and retrieve podcast episodes transcripts.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/devli13/mcp-granola'
If you have feedback or need assistance with the MCP directory API, please join our Discord server
