Kredit Audit · SA SME file

The audit file chain,
on the machine.

Local Windows tools for the SME working-paper ecosystem. Clean the GL, sample it, test journals and revenue — JSON and Excel handed between tools in the client's Kredit folder. Client files are never uploaded.

Local files · JSON and Excel · nothing uploaded

Set up the client
GL Cleaner
Sampler

01 / The goal

Whole ecosystem. Less manual work. Room for judgement.

Combine the audit working papers in the SME space. Remove the manual work — cosmetics, paste, MUS-in-Excel, documentation mechanics — so clerks have time for judgement if the support can be trusted.

Same chain gets the firm ready for local AI that will sit on top of these tools. Files stay local. The interchange is JSON and Excel in the client folder, not a cloud vault.

02 / The chain

One engagement. Tools that hand off.

Each step writes into the client's Kredit folder. The next step reads what the last one left — cleaned GL, sample run, journal selection — so the clerk is not rebuilding the spreadsheet.
  1. Set up the client

    Open the engagement. Year-end in SA date order. IFRS for SMEs when that is the framework. The rest of the chain reads from here.

    Demo · setting-up-client.mp4

  2. GL Cleaner

    Open a messy client GL Excel — page headers, account names sitting in heading rows above the postings. Map the layout. Review what stays against named discards (headers, subtotals, opening balances) and blocked rows. Export cleaned-gl.xlsx and cleaned-gl.interchange.json into the client's Kredit folder. When it is a real ledger, debits equal credits.

    Demo · cleaning-gl.mp4

  3. Sampler

    Reads that JSON. Sections such as Revenue and Expenses. Scope the accounts. Assertion and RMM set the MUS interval from specific materiality and the risk factor. Build debit and credit sides. Reconcile the population to the trial balance — it must agree. Finalise an immutable run with a fingerprint. Export Excel with the testing columns left blank for the file. Method disclosure can show how the sample was computed, so the reviewer can trust the support.

    Demo · sampler.mp4

  4. Journal Sampler

    Performance materiality at line level. Unusual watch-words against routine keywords. Import the journal listing. Completeness: report count versus GL journal count. Exclude the routine (wages, depreciation…). Keep material, year-end, weekend, and unusual. Sign off. Export Excel and JSON into the same folder.

    Demo · journal-sampler.mp4

  5. Revenue WP

    A quiet occurrence paper. Procedures under ISA 240, TB-to-GL recon, credit and debit samples with ticks on the transaction, cut-off, system narrative, audit grid with extrapolation. Nothing leaves the computer.

    Demo · revenue-wp.mp4

  6. Accounting records WP

    Journal-testing paper that imports the Journal Sampler JSON. Preparer → client → materiality → system (who may post; the narrative writes itself) → journals (agreed, authorised, reasonable, query) → ticks and open queries → conclusion.

    Demo · accounting-records-wp.mp4

Local files. SA SME audits. Built by accountants.

Kredit Audit sits on the same desk as Kredit Tools — Windows desktop, cream paper energy, nothing uploaded for a dashboard. The clerk keeps the call. The chain keeps the paper honest.