Skip to main content
Glama
ninetails-io

gnucash-mcp

create_transaction

Create a balanced double-entry transaction in GnuCash with split entries, and validate via duplicate detection or dry-run before writing.

Instructions

Create a new transaction with splits. Splits must balance to zero.

Each split: account (full path, required), amount (required, in transaction currency), quantity (required when account commodity differs from transaction currency), memo (optional), action (optional). amount and quantity are decimal strings (e.g. "94.87") — never raw JSON numbers, which would lose precision on non-dyadic decimals.

FIELD TARGETING — the annotation fields, one job each, in GnuCash-register visibility order:

  • description: the clean name ("Chevron 0090706 Portland"). Always visible.

  • notes: what the purchase WAS, when the description alone doesn't say ("Fuel, road trip to Portland"). Visible in the register's double-line view — this is the annotation humans read. Interpret; don't transcribe.

  • split memo (bank/card leg): the RAW statement line as provenance ("Withdrawal ACH TRAVELERS TYPE: PER INSUR…"). Visible only in expanded split view — evidence, not narrative.

  • split action: the typed KIND of movement, one word. Matters most on investment legs, where desktop convention (and the Advanced Portfolio report) expects "Buy" / "Sell" / "Dividend"; bank legs may use "Wire" / "ATM" / "Interest". Skip it for ordinary spending.

When duplicate detection surfaces candidates (either rejecting the write with status: "rejected" or returning alongside a successful create), duplicates in the response is a newline-separated TSV string, not a list of dicts. Columns::

confidence<TAB>guid<TAB>date<TAB>amount<TAB>cur<TAB>description<TAB>signals

Confidence is HIGH (all three signals match) or MEDIUM (two of three). Signals is a three-char code: position 0 description, position 1 amount (±$1 tolerance), position 2 date (±2 days); D/A/D for match, - for miss.

Args: description: Transaction description. splits: List of split dicts (see above). Omit to auto-fill from the most recent matching-description transaction. transaction_date: ISO date (YYYY-MM-DD). Defaults to today. currency: ISO currency code. Defaults to book's default. notes: What the purchase was (see FIELD TARGETING above). check_duplicates: Run duplicate detection. Default True. force_create: Create even if HIGH-confidence duplicates found. dry_run: Validate + dupe check only; don't write.

Input Schema

TableJSON Schema
NameRequiredDescriptionDefault
notesNo
splitsNo
dry_runNo
currencyNo
descriptionYes
force_createNo
check_duplicatesNo
transaction_dateNo

Output Schema

TableJSON Schema
NameRequiredDescriptionDefault
resultYes
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

No annotations are provided, so the description carries full burden and it delivers richly: discloses the duplicate-detection behavior (rejected status or returned alongside success), the exact TSV format of the duplicates response with column layout and signal coding, force_create and dry_run semantics, and the decimal-string requirement explaining why raw JSON numbers lose precision. This is outstanding behavioral disclosure.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Dense but well-structured with clear section headers (FIELD TARGETING, duplicate-detection block, Args list). Every sentence earns its place given the tool's real complexity. Slightly long, but the length is justified by the genuine ambiguity around field targeting and duplicate handling that would otherwise confound an agent.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a tool with 8 params, 0% schema coverage, no annotations, but a complex duplicate-detection response contract, this description is exceptionally complete. It covers split balancing, field semantics per GnuCash register convention, response format details, and all behavioral flags. Return-value handling is disclosed even though an output schema exists.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters5/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

Schema description coverage is 0%, so the description must compensate entirely — and it does comprehensively. Every parameter is explained: splits structure with each sub-field (account, amount, quantity, memo, action), date format, currency default behavior, notes interpretation guidance, and omissions like splits auto-filling from most-recent matching transaction. Also details the decimal-string formatting requirement that the schema barely hints at.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

Clear verb+resource: 'Create a new transaction with splits. Splits must balance to zero.' Distinguishes from siblings like create_transactions (plural, batch) and create_transaction_from_scheduled (derived from template). The scope and invariants are explicit.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

Extremely detailed FIELD TARGETING section explains exactly what goes in description vs notes vs split memo vs action, with concrete examples for each. This is exceptional when-to-use guidance that prevents common misuses (e.g. putting raw statement text in description instead of the clean name, or filling action for ordinary spending). Contrasts with create_transactions implicitly via 'create over a single transaction'.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Install Server

Other Tools

Latest Blog Posts

MCP directory API

We provide all the information about MCP servers via our MCP API.

curl -X GET 'https://glama.ai/api/mcp/v1/servers/ninetails-io/gnucash-mcp'

If you have feedback or need assistance with the MCP directory API, please join our Discord server