Why we do this: Workspace settings

Workspace settings control provider/privacy defaults, shared risk configurations, holidays, and the versioned firm sampling table.

What you do
Review configuration, make any justified edits, provide change reasons, and inspect version history.
What the system does
Saves shared settings and sampling-table versions in the database; credentials remain in the server environment.
What you decide
Approve privacy choices, defaults, rounding, and changes to the firm table.
Output
Workspace configuration, table history, named profiles, and custom rules.

Common mistakes

  • Changing the firm table without assessing its effect on future runs.
  • Assuming an API key is stored in the database or exposed to the browser.

Relevant guidance: Firm policy

Workspace configuration

Settings

Preferences save to the database. API keys stay only in the server-side environment.

Provider · MockMasking · onRounding · up0 holidays

Step 1

AI provider and privacy

Model calls are advisory; they do not calculate materiality or choose the sample.

Default provider

?

Secret contents are never sent to this page.

Connection test

Checks the selected provider without sending engagement ledger data.

?
Stable placeholders are used for counterparties, references, and accounts in cloud requests.

Step 2

Materiality defaults ?

These are starting defaults for new engagements; each engagement can override them. Auditors select the benchmark and justify every locked calculation.

Benchmark percentages

Thresholds

Performance materiality must be between 50% and 75%. Clearly trivial must be between 1% and 20%.

Step 3

Sampling factor rounding ?

The selected rule is recorded in each future sample run.

Rounding down can produce a smaller sample than rounding up. Choose the rule that matches your firm's methodology.

Step 4

Must-sample transaction categories

These transaction names and account descriptions are always treated as high-priority audit selections, even without a large amount or a strong reason-code signal.

The default list covers common fraud-prone expense categories such as professional fees, legal fees, audit fees, subscriptions, marketing, and donations.

Step 4

Sampling table

Active version loading

Export as CSV
Changing this table changes the suggested sample sizes for future runs. Existing sample runs keep their saved table version and row.

Factor table

0 rows
Needed to save, reset, restore or import. Write at least 10 characters.0 / 1000
Version history (0)

No saved versions to show yet.

Step 5

Public holidays

One ISO date (YYYY-MM-DD) per line; used by weekend/holiday risk rules.

0 dates

Step 6

Legacy risk rule weights

Retained for compatibility with saved runs; the new sample uses reason-code configuration.

MaterialityJournal entryMissing evidencePeriod endCurrent total: 90

All changes saved