What it does
Answers one question before a period is closed: are the books ready, or not yet?
The same dimensions, the same tests and the same page every close, so a scorecard made on day 3 and one made on day 5 can be laid side by side.
It uses the month-end checks and the journal screen when those skills are installed, and runs reduced tests when they are not.
Six dimensions
- Bank and cash agree
- Everything is in
- Tax and payroll tie
- Movements are explained
- The balance sheet makes sense
- Evidence and trail
Four verdicts
- Ready to close
- Nothing in the rubric stands in the way. It does not mean the figures are correct.
- Ready with notes
- Not yet
- Limited read
- Only one trial balance was given. A limited read is never Ready to close.
What the script works out
From two trial balances, the bundled script tb_compare.py works out:
- Whether the trial balance balances
- Movements above 10 percent and above a floor of 0,5 percent of income to date
- Income and expense activity against the monthly run rate
- Holding accounts with a balance, and whether it grew
- Balances on the wrong side
- Bank, stock, accrual and depreciation balances that did not move
- Accounts that went quiet
- New and emptied accounts
- Retained earnings that moved
Two kinds of finding, always labelled
- Worked out
- A figure compared with a figure.
- Read from the balances
- An inference, worded as a question to you.
The scorecard
Every run writes the same sections, in this order.
- The verdict
- Each dimension, with its status and open items
- Findings, with weight, account, finding, kind and next step
- Movements, each Explained, Likely reason or Unexplained
- What was not assessed
- The path to ready
- Questions
What it reads
Trial balances at the period end and at the prior period end, as CSV or Excel, or a ledger export with opening balances.
With only one trial balance, the verdict is Limited read.
Try “Is my June close ready? Here are the trial balances for 31 May and 30 June.”
Connected to Cybooks
Reads the live trial balance at both dates, and explains each flagged movement from the ledger lines behind it.
It places the checks Cybooks already ran and the month’s open to-dos into the dimensions.
What you need
- Claude, with skills
- Anthropic offers skills on its Free, Pro, Max, Ultra, Team and Enterprise plans. To add this one, open Customize › Skills, press +, then Create skill › Upload a skill, and pick its file.
- Code execution
- Turn on Code execution and file creation in Settings › Capabilities, or use Claude Code. Every figure is computed by code, not estimated by the model.
- On its own
- CSV or Excel exports. It sees only the files you attach.
- Connected
- The Cybooks connector at https://app.cybooks.com.cy/mcp/cybooks, signed in. It reads with your own permissions: view reports, view the books review and view the dashboard. Two optional steps also need manage the period lock and view manual journals. The connector is on Premium.
How it was tested
Tested with 31 script tests, a real six-month general ledger export, and two rounds of end-to-end runs by a fresh Claude agent with only the skill file and the raw export. The changes from the second round and from a code review (13 issues) are covered by the script tests, not yet by a fresh end-to-end run. Not yet tested: a run inside the Claude apps, and the connected path against the live Cybooks connector.
Claude is a trademark of Anthropic. Cybooks is not affiliated with Anthropic; no endorsement implied.