Click on "Deploy 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., "@Lunch Money MCP Server show me my transactions for July 2024"
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.
A Model Context Protocol (MCP) server for the Lunch Money API v2, designed with minimal response sizes to prevent context window bloat.
Optimized responses: Concise, formatted output to minimize token usage
Simple authentication: Uses environment variable for API token
Type-safe: Built with modern Python type hints
Easy to extend: Add more endpoints one at a time
Related MCP server: LunchMoney MCP Server
add_numbers - Helper tool for arithmetic operations
get_current_user - Get information about the authenticated user (GET /me)
get_transaction - Get details about a specific transaction by ID (GET /transactions/{id})
get_transactions - List transactions for a date range (GET /transactions)
Clone this repository:
git clone <your-repo-url>
cd lunchmoney-mcp-miniInstall dependencies using uv:
uv syncLog in to Lunch Money
Go to the Developers page
Create a new API token or use an existing one
export LUNCHMONEY_API_TOKEN="your-api-token-here"Or create a .env file (not committed to git):
LUNCHMONEY_API_TOKEN=your-api-token-hereAdd to your Claude Desktop configuration file:
macOS: ~/Library/Application Support/Claude/claude_desktop_config.json
Windows: %APPDATA%\Claude\claude_desktop_config.json
{
"mcpServers": {
"lunchmoney-mini": {
"command": "uv",
"args": [
"--directory",
"/absolute/path/to/lunchmoney-mcp-mini",
"run",
"lunchmoney_mcp_mini/main.py"
],
"env": {
"LUNCHMONEY_API_TOKEN": "your-api-token-here"
}
}
}
}# Make sure LUNCHMONEY_API_TOKEN is set
uv run lunchmoney_mcp_mini/main.pyadd_numbersHelper tool for performing arithmetic operations with precise decimal handling to avoid floating-point precision issues.
Parameters:
numbers (required): List of numbers to add together. Can include negative values for subtraction.
Returns:
sum: Sum rounded to 2 decimal places
input_count: Number of values provided
Example output:
{
"sum": 123.45,
"input_count": 3
}get_current_userGet details about the authenticated Lunch Money user.
Returns:
name: User's full name
email: User's email address
user_id: Unique user identifier
account_id: Unique account identifier
budget_name: Name of the budget
primary_currency: Primary currency code (e.g., 'usd')
api_key_label: Label for the API key being used
Example output:
{
"name": "John Doe",
"email": "john@example.com",
"user_id": 12345,
"account_id": 67890,
"budget_name": "Family budget",
"primary_currency": "usd",
"api_key_label": "Development key"
}get_transactionGet full details about a specific transaction by its ID.
Parameters:
transaction_id (required): ID of the transaction to retrieve
Returns: Complete transaction object with all available fields including:
Core data: id, date, amount, currency, payee, original_name
Category/accounts: category_id, manual_account_id, plaid_account_id, recurring_id
Metadata: plaid_metadata, custom_metadata, files (if any)
Grouping/splitting: is_split_parent, split_parent_id, is_group_parent, group_parent_id, children
Timestamps: created_at, updated_at
Status: status, is_pending, source, external_id, tag_ids, notes
Example output:
{
"id": 2112150655,
"date": "2024-07-28",
"amount": -45.50,
"currency": "USD",
"payee": "Whole Foods",
"original_name": "WHOLE FOODS #1234",
"category_id": 82,
"status": "reviewed",
"is_pending": false,
"created_at": "2024-07-28T12:34:56.789Z",
"updated_at": "2024-07-28T12:34:56.789Z"
}get_transactionsList transactions within a specified date range.
Parameters:
start_date (required): Start date in YYYY-MM-DD format
end_date (optional): End date in YYYY-MM-DD format. Defaults to last day of start_date's month
category_id (optional): Filter by category ID
tag_id (optional): Filter by tag ID
status (optional): Filter by status ("reviewed", "unreviewed", "delete_pending")
is_pending (optional): Filter by pending status
manual_account_id (optional): Filter by manual account ID
plaid_account_id (optional): Filter by plaid account ID
recurring_id (optional): Filter by recurring item ID
include_pending (optional): Include pending transactions
limit (optional): Maximum number of transactions (1-2000, default 100)
offset (optional): Pagination offset
include_aggregates (optional): If True, calculates totals per category for full date range (respects all filters)
Returns:
transactions: Array of transaction objects
has_more: Boolean indicating if more transactions are available
aggregates (optional): Category totals and counts when include_aggregates=True
Transaction fields:
id: Transaction ID
date: Transaction date (YYYY-MM-DD)
amount: Transaction amount (numeric string)
payee: Payee name
category_id: Category ID
status: Transaction status
is_pending: Pending status
Aggregates fields (when include_aggregates=True):
by_category: Array sorted by total_amount descending, each with:
category_id: Category ID (or null for uncategorized)
category_name: Category name
count: Number of transactions in this category
total_amount: Sum of transaction amounts (numeric string)
total_count: Total number of transactions
total_amount: Sum of all transaction amounts (numeric string)
Example output (without aggregates):
{
"transactions": [
{
"id": 2112150655,
"date": "2024-07-28",
"amount": "1250.8400",
"payee": "Paycheck",
"category_id": 88,
"status": "reviewed",
"is_pending": false
}
],
"has_more": false
}Example output (with aggregates):
{
"transactions": [...],
"has_more": false,
"aggregates": {
"by_category": [
{"category_id": 88, "category_name": "Rent", "count": 2, "total_amount": "2500.00"},
{"category_id": 82, "category_name": "Groceries", "count": 5, "total_amount": "245.50"},
{"category_id": null, "category_name": "Uncategorized", "count": 3, "total_amount": "45.00"}
],
"total_count": 10,
"total_amount": "2790.50"
}
}This MCP server is intentionally designed to return minimal, focused responses to avoid filling up the context window. Each tool:
Returns only essential information
Uses concise formatting
Avoids verbose JSON dumps
Provides human-readable output
This server uses:
FastMCP: A high-level Python framework for building MCP servers
requests-openapi: Automatically generates API client from OpenAPI spec
OpenAPI 3.0 spec: Ensures type safety and accurate API calls
The combination of FastMCP and requests-openapi means:
Less boilerplate code
Automatic request/response validation
Easy to add new endpoints from the spec
Type-safe API calls
MIT
Helper tool for adding numbers together.
LLMs should use this tool for arithmetic operations to avoid calculation errors. This is especially useful for summing expenses, calculating totals, or performing any arithmetic where precision matters.
Args: numbers: List of numbers to add together. Can include negative values for subtraction.
Returns: Dictionary with the sum rounded to 2 decimal places to avoid floating-point precision issues.
| Name | Required | Description | Default |
|---|---|---|---|
| numbers | Yes |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations, but description details return format (rounded to 2 decimal places) and supports negative numbers for subtraction. Sufficiently transparent for a simple addition tool.
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?
Concise with clear Args and Returns sections. Every sentence is useful, no fluff.
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 simple input schema and output schema mentioned, the description fully covers behavior, parameter usage, and return format. Complete for the tool's purpose.
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 coverage is 0%, but description explains the numbers parameter as 'List of numbers to add together' and mentions negative values for subtraction, adding value beyond 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 'Helper tool for adding numbers together' and provides concrete use cases like summing expenses and totals, distinguishing it from unrelated sibling tools.
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?
Explicitly recommends using this tool for arithmetic operations to avoid errors, with examples. Does not explicitly state when not to use, but context makes it clear it's for arithmetic only.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get all accounts (both manual and Plaid-synced).
Returns a combined list of all accounts in the user's budget, including:
Manual accounts: manually-managed asset/liability accounts
Plaid accounts: accounts synced with financial institutions
Each account includes minimal information: id, name, type, balance, currency, status, and account_type (to distinguish between 'manual' and 'plaid' accounts).
Useful for understanding which accounts are available and their current balances.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
The description discloses that the tool returns a combined list with minimal information for each account, including specific fields. Since there are no annotations, it adequately covers the read-only nature and output structure, though it could mention if there are any limitations like pagination.
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 concise with three sentences, front-loaded with the main action, and each sentence adds meaningful information without redundancy.
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 presence of an output schema, the description adequately covers the tool's purpose and return value. It lists the main fields but could mention error conditions or prerequisites for completeness, though the simplicity of the tool mitigates this need.
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?
There are zero parameters, and the baseline score for such cases is 4. The description adds value by detailing the types of accounts included and output fields, which goes beyond what the empty 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 'Get all accounts' and specifies it includes both manual and Plaid-synced accounts, distinguishing it from sibling tools that handle other resources like categories and transactions.
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 mentions 'Useful for understanding which accounts are available and their current balances,' which implies a usage context, but it does not explicitly state when to use this tool over alternatives or when not to use it.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Return a list of category names.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
| result | Yes |
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries the full burden. It simply states the outcome without disclosing side effects, permissions, or whether the list is exhaustive. The agent cannot infer read-only behavior or other traits.
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 concise sentence with no unnecessary words. It is front-loaded and effectively communicates the core function without redundancy.
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 simplicity (0 parameters, output schema exists), the description is minimally adequate. However, it lacks context about the scope of categories (all user's? all system?), ordering, or how it relates to sibling tools, leaving some ambiguity.
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 tool has 0 parameters, so baseline is 4. The schema coverage is 100%, and the description adds no parameter details, which is acceptable since no parameters exist.
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 verb 'Return' and the resource 'list of category names', making the tool's purpose evident. It distinguishes itself from sibling tools like 'get_transactions' by focusing on categories, though it does not explicitly differentiate.
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 no guidance on when to use this tool versus alternatives. There is no mention of context, prerequisites, or exclusions, leaving the agent to infer usage.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get details about the current Lunch Money user.
Returns user information including:
name: User's full name
email: User's email address
user_id: Unique user identifier
account_id: Unique account identifier
budget_name: Name of the budget
primary_currency: Primary currency code (e.g., 'usd')
api_key_label: Label for the API key being used (or null)
Useful for verifying authentication and understanding the account context.
| Name | Required | Description | Default |
|---|---|---|---|
No parameters | |||
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description carries full burden. It implies a read-only operation by stating it returns user information, but it does not explicitly confirm no side effects or mention authentication requirements. The listed return fields add some transparency.
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: first states the core function, second lists key fields and use cases. No unnecessary words; front-loaded with purpose.
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?
The tool is simple with no parameters and an output schema (presence noted). The description covers the return fields and usage, and sibling tools are distinct. Complete for the tool's complexity.
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 tool has no parameters and schema coverage is 100%, so the description's role is minimal. It adds value by listing return fields, though this pertains to output rather than parameters. Baseline 4 applies for zero-parameter tools.
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 tool retrieves details about the current Lunch Money user, listing specific fields like name, email, user_id. It distinguishes itself from sibling tools (e.g., get_categories, get_transactions) which target different resources.
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 notes the tool is 'useful for verifying authentication and understanding the account context,' providing clear usage context. It does not explicitly exclude alternatives, but the context is sufficient for an agent to decide when to use it.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Get details about a specific transaction.
Retrieves the full details of a single transaction by its ID, including:
Core data: id, date, amount, currency, payee, original_name
Category: category name (and category_id for reference)
Accounts: manual_account_id, plaid_account_id, recurring_id
Metadata: plaid_metadata, custom_metadata, files (if available)
Grouping/splitting: is_split_parent, split_parent_id, is_group_parent, group_parent_id, children
Timestamps: created_at, updated_at
Status: status, is_pending, source, external_id, tag_ids, notes
Args: transaction_id: ID of the transaction to retrieve
Returns: Full transaction object with all available fields
| Name | Required | Description | Default |
|---|---|---|---|
| transaction_id | Yes |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
No annotations are provided, so the description carries full burden. It lists all fields returned, which adds clarity about the output. The phrase 'Get details' implicitly indicates a read-only, non-destructive operation, but an explicit statement about safety (e.g., 'This is a safe read operation') is missing.
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 well-structured with bullet points, but it is verbose, listing all return fields. Since an output schema exists, this redundancy reduces conciseness. The opening sentence is good, but the detailed list could be omitted.
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?
The tool is simple (1 parameter, output schema available). The description covers the parameter, explains what is returned (though redundant), and provides context. It is complete for this tool, but could explicitly state it is a read-only operation.
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 input schema has 0% description coverage, but the description explicitly documents the parameter as 'transaction_id: ID of the transaction to retrieve,' providing meaning beyond the schema's type-only definition. With a single parameter, this is clear and sufficient.
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 it gets details of a specific transaction, using a specific verb+resource ('Get details about a specific transaction'). It distinguishes from the sibling tool 'get_transactions' which retrieves multiple transactions.
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 use for a single transaction via 'specific transaction' and mentions the transaction_id parameter, but does not explicitly state when to use this tool versus alternatives like get_transactions, 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.
Get transactions for a date range.
This is a paginated tool and you MUST consider that not all transactions
may be returned. The has_more return value will tell you if pagniation
should continue. If has_more is true, ask yourself if you need to make
another request to properly answer the user's query.
include_aggregates returns aggregates for all pages.
Args: start_date: Start date in YYYY-MM-DD format (required) end_date: End date in YYYY-MM-DD format (defaults to last day of start_date's month) filter_category_name: Filter by category name (e.g., "Groceries", "Dining Out") filter_tag_id: Filter by tag ID filter_status: Filter by transaction status (reviewed, unreviewed, delete_pending) filter_is_pending: Filter by pending status filter_manual_account_id: Filter by manual account ID filter_plaid_account_id: Filter by plaid account ID filter_recurring_id: Filter by recurring item ID include_pending: Include pending transactions (ignored if is_pending is set) limit: Maximum number of transactions to return (1-100, default 100) offset: Pagination offset include_aggregates: If True, calculates totals per category for full date range (respects all filters, except pagination)
Returns: Structured JSON where transactions include category names instead of IDs, has_more pagination flag, and optionally category aggregates. When has_more is true, next_offset and next_limit are provided for easy pagination.
| Name | Required | Description | Default |
|---|---|---|---|
| limit | No | ||
| offset | No | ||
| end_date | No | ||
| start_date | Yes | ||
| filter_status | No | ||
| filter_tag_id | No | ||
| include_pending | No | ||
| filter_is_pending | No | ||
| include_aggregates | No | ||
| filter_recurring_id | No | ||
| filter_category_name | No | ||
| filter_plaid_account_id | No | ||
| filter_manual_account_id | No |
| Name | Required | Description |
|---|---|---|
No output parameters | ||
Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?
With no annotations, the description thoroughly discloses pagination, return structure (transactions with category names), and aggregate behavior. It provides all necessary behavioral traits.
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?
Well-structured with a clear purpose, pagination warning, parameter list, and return details. Slightly verbose but every sentence is informative. Front-loaded with essential info.
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?
Fully covers the tool's complexity: 13 params, pagination, aggregates, return format. With no annotations and an output schema present, the description is sufficient for correct invocation.
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?
Despite 0% schema coverage, the description explains all 13 parameters with formats, defaults, and dependencies (e.g., 'include_pending ignored if is_pending set'). Adds significant value beyond 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 'Get transactions for a date range' and details the functionality, distinguishing it from sibling tool 'get_transaction'.
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?
Explicitly instructs about pagination, how to handle 'has_more', and when to make additional requests. Also clarifies 'include_aggregates' behavior.
Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
v0.5.0add_numbersget_accountsget_categoriesget_current_userget_transactionget_transactionsScored across 6 tools
Each tool targets a distinct resource or action: user info, categories, arithmetic, transaction list, single transaction, and accounts. No overlap or ambiguity.
All tools use a consistent verb_noun pattern in snake_case (e.g., get_current_user, get_transactions, add_numbers). No mixing of conventions.
6 tools is well-scoped for a personal finance MCP server covering user, categories, transactions, and accounts with a helper arithmetic tool. Not too many or too few.
The server is heavily read-only, missing write operations for transactions, categories, and accounts. The arithmetic helper is out of domain, and there are no create, update, or delete tools.
Read-only Lunch Money accounts, transactions, categories and budgets. Unofficial connector.
Query your real net worth, spending, transactions, budgets and portfolio from any MCP client.
SmartMoney77 MCP v0.6.0 — 14 public tools that turn financial questions into exact numbers and citable links. New: historical_investment_return and compare_investments, which compute "what if I had invested" results from real yearly price data. Also compound interest, FIRE number, credit-card payoff, emergency fund, inflation, latte factor, investment fees, cost of waiting, plus discovery/deep-link/share-pack tools for a catalog of calculators in 6 languages (he/en/ar/es/pt/in). Public, no login. Endpoint: https://smartmoney77.com/mcp
Log, query, and edit expenses, budgets, and accounts in Manilo from any MCP-compatible AI assistant.
