Use case
How to get finance-grade records for AI spend
A logging table that gets edited when someone finds a mistake is not a record finance can sign off on.
The append-only ledger is live today; sealing a month is not, it is gated behind a flag not yet enabled in production. Live half: every call is written once and a correction is always a new row, never an edit to history. Designed half: a month-close state machine, open, ready, closed, reconciled, that makes a closed period immutable.
How Spendline does this
Two properties make a ledger finance-grade rather than just a log: nothing already written is ever mutated or deleted, and a month is designed to be sealed. Once a month moves to closed, no new call is meant to post to it and no existing entry is meant to change; a mistake discovered later is corrected with a new adjustment row referencing the original, so the full history, including the mistake and its correction, stays visible. Actually moving a month to closed, or reopening or reconciling one, is gated behind a flag not yet enabled in production; the append-only guarantee on the ledger itself does not depend on that flag.
Why it matters
An auditor, or a finance lead preparing a board number, needs to trust that a number they signed off on last month cannot silently change this month. A mutable spend table cannot make that guarantee; an append-only ledger with an immutable close can.
Frequently asked questions
What if a cost was attributed wrong and the month is already closed?
A ledger_adjustments entry is added as a new row correcting it. The original entry is never edited or deleted, so the audit trail shows both the original and the correction.
Who can close a month?
By design, an owner-authorized action, not something available to every API key, since closing, reopening, or reconciling a period changes an immutable financial record. Today the close endpoints return a 503 while disabled in production, so this describes who will be authorized once it is turned on, not something usable right now.
See where this stands in your own setup
This page describes the mechanism. The 5 minute assessment scores your own attribution, enforcement, and reconciliation setup, so you know exactly which of these problems you actually have today.
Run the 5 minute assessment