Dr Audit

Audit Evidence Package

Generate an audit-support evidence package over FinanceOS data — the control checks that this data surface can actually evidence, packaged for management and auditors.

Creates both a PDF report (for management) and an Excel evidence workbook (for the audit trail).

Honest scope — read first. This is not a SOX certification. In scope are the data-evidencable control families: completeness & period integrity, consistency/reconciliation, account-mapping integrity, and substantive sampling — each backed by a real query this skill can run. Access control, change management, and IT general controls are out of scope: the FinanceOS MCP surface has no audit-log, access-history, or user-activity endpoint, so those control families require system-administration evidence outside this tool's reach. Never present a control result this skill cannot substantiate with a tool call — the evidence workbook carries a mandatory "Out of scope — requires external evidence" sheet so no reader mistakes the package for full SOX coverage.

Excel Context — Routing Preamble

Before any data pull, establish whether this skill is running in a live Excel context (Claude for Excel with the Datarails Add-In loaded) and route accordingly.

Detect — never infer from the user's wording. A sheet list containing __dr_agent means the add-in is loaded. Confirm with the agent.get_session probe, which you run by executing Office.js through the execute_office_js tool (see the Excel Context Contract in CLAUDE.md, §Transport) — it is not an MCP tool and has no MCP equivalent. A failed probe is a normal detection result, not an error: it means "no bridge here", which is the expected outcome in Claude Code. Do not surface it, do not retry it, and do not apply this skill's connection-error or Connectors-UI guidance to it — that guidance is about datarails-finance-os connector calls only. A successful probe means Excel context, on either transport. The bridge serves two add-in tracks and their session payloads differ: Flex (Office.js task pane) exposes isLoggedIn; the COM desktop add-in — the majority of live workbooks — exposes isConnected instead, and isConnected: false is not a login failure, an error, or a reason to stop or send the user anywhere. It merely means the workbook isn't connected to a Datarails file, which matters only to drilldown_* / create_dynamic_range (the bridge skill gates those itself). Only Flex's explicit isLoggedIn: false means sign-in is needed.

Route by the target of the request, not by whether a workbook is open.

• Org / server data — which tables, models and fields exist, aggregations, raw rows, distinct values, metrics, profiling — always the datarails-finance-os MCP connector, even in Excel. The bridge cannot answer these.
• Workbook actions — refresh, drill a cell, insert a DR function, read what a cell returns, publish, submit — always the add-in bridge. Never a native Excel recalc (calculate(), F9): it does not pull Datarails data and silently yields stale values.

In a live Excel context this skill cannot produce its file deliverable. Its generation steps depend on the Bash tool, which that surface does not provide. Say so plainly and offer the real alternatives — a scoped answer in chat, or re-running this skill from Claude Code where file output works. Never improvise another route to a file, never hand back a partial artifact, and never silently substitute a different deliverable: writing into someone's live workbook instead of giving them the file they asked for is a different and irreversible outcome, not a smaller version of the same one.

If you do write DR formulas into the workbook, writing and refreshing are one atomic step. Write to a new sheet, fire refresh_selected_cells_ribbon scoped to that range — a new-sheet block is one contiguous range, so one scoped call covers any cell count — then read the range back. refresh_ribbon is not the tool for this: it repulls every DR cell in the file and can silently move numbers elsewhere in the user's model. It is reserved for the one case the scoped command can't cover — scattered inserts across multiple sheets, per excel-context__internal's refresh-after-insert rule — and even then only with the user's explicit OK, after snapshotting the DR ranges you can bound, reporting each changed cell in them before → after with the compared ranges named, and saying plainly that cells beyond them may also have updated. If the user declines the whole-workbook refresh, fall back to scoped refresh_selected_cells_ribbon calls sheet-by-sheet — slower, but nothing outside the written cells moves. A freshly written DR formula reads Missing / Loading… / #BUSY! until an agent refresh lands, so never quote a value you have not read back after a successful refresh, and never present a figure fetched from the MCP connector as though it were the cell's value. /dr-get-formula is the full authority for DR.GET workbooks.

If the user asks you to elaborate on a DR-backed figure — "explain", "break down", "what's driving this", "why is X" — and the figures in scope are DR formula cells, offer the add-in's drill-down instead of silently re-deriving the number through the MCP connector. A drill resolves the exact filters behind that cell; a hand-rebuilt query only approximates them.

Arguments

Control Checks (data-evidencable only)

Each check maps to tools this skill can actually call — nothing goes in the evidence package without a query behind it:

1. Completeness & period integrity. Discover the date field's range and confirm every expected period in the audited quarter/year is present (distinct values of the reporting-month field). Then verify integrity with two grouped calls over the audited window — by scenario and by period: first apply the item-4 defensive filter — keep only rows in which every requested dimension key is present, preserving genuine [null] groups (a stale cached response can carry legacy roll-up rows that omit keys and equal the whole total, faking a mismatch) — then sum those rows (including [null]) to get that slicing's total; responses carry one row per group and no total row in the data (Data-scope preamble, item 4). The responses must be complete: on truncated: true, narrow (e.g. split the window into sub-ranges) and stitch — never sum a prefix, which would compare two truncated prefixes and report a false pass or fail. Do not substitute the top-level totals field here: the engine computes it independently of the grouping, so both slicings would return the same number by construction and the check would pass vacuously. The two totals must equal each other, since both describe the same window sliced two ways. Tools: start_distinct_values_by_alias/_by_id → poll get_distinct_values_result_by_alias/_by_id, and start_aggregation_by_alias/_by_id → poll get_aggregation_result_by_alias/_by_id (async-fetch pattern).

2. Consistency / reconciliation. The reconciliation control is the /dr-reconcile skill's four independent-source checks — cross-endpoint agreement, balance-sheet identity, cross-grain roll-up, and scenario/period integrity. That skill's SKILL.md is the single source of the method (query shapes, tolerance, pass/fail rules) — but its native run is year-scoped (--year is its only window argument), while this audit is quarter-scoped. So do not delegate a bare full-year run: apply the four checks' method with the date filter narrowed to the audited window (--year + --quarter, the same advanced date-range filter used by every other family in this evidence package), and record in the evidence which window each check actually covered. A full-year reconciliation pass does not evidence a Q-scoped control — a quarter-local inconsistency can net out over the year.

3. Account-mapping integrity. Deliberately reuses the roll-up mechanics from the reconciliation control's Check 3 — but for a different verdict: not "does the pipeline roll up consistently" (that pass/fail belongs to check family 2 above) but "which accounts are unmapped". Two aggregates over the same scope — dimensions=[<parent_level>] and dimensions=[<parent_level>, <child_level>], both SUM(<amount>), with the same parameters, tolerance, and audited-window date filter as family 2's roll-up check (year + quarter — never the full year) so both comparisons cover the same period. Sharing the scope aligns the comparison; it does not make the totals agree — a mismatch is exactly the finding. For each parent bucket, the sum of its child rows — including the [null] bucket — must equal the parent's own row to the cent.

   Then name the accounts — the two aggregates above cannot. They return bucket totals, so the presence of a [null] child bucket tells you a parent has unmapped rows, not which accounts carry them. An exception log that names no account is not auditable evidence.

   Trigger on row presence, never on a non-zero amount. Unmapped rows with offsetting positive and negative amounts net to zero, so a [null] bucket summing to 0.00 can still hold unmapped accounts — gating on the amount would silently drop those exceptions from the evidence package. Add a row count to the family-3 aggregate alongside the sum — metrics=[{SUM of <amount>}, {COUNT of <non_null_row_identifier>}] — and treat count > 0 as the trigger. SUM(<amount>) reports the net unmapped amount only; it never decides whether to look.

   Choosing <non_null_row_identifier> is load-bearing: COUNT skips nulls, so it must be a field the discovered schema populates on every source row (a system row id, or the reporting-date field). It must not be the account identifier or the child level — those are exactly the fields that are null on the rows being hunted, so counting them can return 0 while unmapped rows exist, reintroducing the miss this rule prevents. It must also differ from <amount>, since one call may carry only one aggregation per field. Verify non-nullness from the field's profile (null rate 0) before relying on it; if no such field exists, fall back to the bounded row-level query below and treat any returned row as the trigger.

   For each parent whose [null] child bucket contains rows, run one follow-up query under the same period + scenario filters:

   • Preferred — re-aggregate that parent with the account identifier added as a dimension: dimensions=[<parent_level>, <child_level>, <account_id_or_name_field>], SUM(<amount>), filtered to that parent. Read the rows whose <child_level> is the [null] bucket — those accounts, with their amounts, are the exception-log entries.
   • When row-level proof is wanted — get_data_by_alias / get_data_by_id with select on the account + amount + date columns and filters scoping that parent, the audited window, and the scenario (the is null advanced condition isolates the unmapped child set). Respect the 500-row cap and the truncated envelope, per family 4.

   The exception log records account identifier, parent bucket, amount, and the window. The roll-up pass/fail itself is reported once, under family 2 — this family reports only the exception list.

4. Substantive sampling. For the material buckets (largest by absolute amount in the audited window), pull row-level detail via get_data_by_alias / get_data_by_id with select on the load-bearing columns and filters scoping bucket + period, so a human auditor can trace reported figures to source line items. Respect the 500-row cap — sample per bucket, never attempt a full extract. If a response arrives with "truncated": true, the returned rows are an incomplete prefix — narrow the query per its guidance (more filters / fewer columns / lower limit+offset paging) and re-fetch; never present the prefix as a complete sample.

Out of scope — requires external evidence: access control, change management, and IT general controls. No tool available to this skill can observe user access, permission grants, or change history — do not test, score, or opine on these families; list them on the out-of-scope sheet instead.

All of these checks aggregate live data. Run this discovery before any check that queries or aggregates:

Async fetch — aggregations and distinct values run as start → poll. start_aggregation_by_id/_by_alias and start_distinct_values_by_id/_by_alias take the same arguments as the retired blocking calls (dimensions/metrics/filters; table id + field id, or alias + field alias) and return immediately with {"status": "pending", "handle": {...}}. Echo that handle back verbatim to the matching get_aggregation_result_by_* / get_distinct_values_result_by_* tool: a {"status": "running", "retry_after_seconds": N} response means poll again with the same handle after ~N seconds (≈5s) — it is not an error, and large jobs may take several polls; when ready, the result arrives in the familiar shape (for distinct values, pass limit to the result tool). An expired/unknown-handle error means restart with the start_* tool. Transitional fallback: if the start_* tools aren't available on the connector (older server), the blocking twins get_aggregated_data_by_* / get_distinct_values_by_* still work with the same arguments.

Data-scope discovery — run before any aggregate (reuse anything already discovered this conversation).

1. Scenario domain. Pull distinct values of the scenario field (start_distinct_values_by_alias/_by_id → poll the matching result tool) — never assume a scenario name exists (Budget frequently doesn't; many orgs carry only {Actuals, Forecast}). For budget/plan questions, if no budget-like scenario exists, look for a planning-version-like field (alias/name matching /plan|version|cycle|budget/i) and use its versions as the plan side; if neither exists, say so and offer a comparison across the scenarios that do exist.
2. Account grain. Pull distinct values of each account-hierarchy level field (L0/L1/L2-like). Use the level whose values partition P&L flows into revenue/COGS/opex-like buckets — on many orgs the top level is the balance-sheet equation (ASSET/LIABILITY/EQUITY/INCOME) and P&L line items live one level deeper. For P&L work, scope to P&L flows and exclude balance-sheet buckets; never present asset/liability/equity totals as revenue or expenses.
3. Period scope. Discover the date field's range (distinct values of the reporting-month field, or MIN and MAX in two separate calls — one aggregation per field per call). Default every P&L question to the latest complete fiscal year (or trailing 12 closed months) — never an unscoped all-time total: financials tables are multi-year cumulative and mix balance-sheet stock with P&L flow. Label every output with the period + scenario it covers.
4. Reading GROUP BY responses. Each response returns exactly one row per requested group — no subtotal rows and no grand-total row mixed into the data list; grand totals arrive in a separate top-level totals field beside the rows ({"data": [...], "totals": {...}}), computed across all groups, not just the returned prefix. For a grand total, read totals — never sum the rows when the response carries truncated: true (summing the returned prefix silently under-counts; dev repro: 474 of 31,455 rows summed to 21% of the true total). totals combines the per-group results rather than re-scanning the rows, so it is exact exactly when the aggregation is decomposable: SUM (sum of the group sums), COUNT (sum of the group counts), MIN, and MAX. It is WRONG for AVG (unweighted mean of the group averages) and COUNT_UNIQUE (sum of the per-group distinct counts, so a value recurring across groups is counted once per group) — true average = SUM total ÷ COUNT total (two calls: a field may be aggregated at most once per request); true distinct count = the distinct-values tools. Treat every aggregation type not named exact above — UNIQUE_VALUES included, whose cross-group de-duplication is unverified (the COUNT_UNIQUE behaviour above is evidence the engine may not de-duplicate across groups at all) — as not decomposable: derive it from complete rows or the distinct-values tools, never from totals. totals is absent on dimension-less aggregations (the single returned row IS the total) and may be absent on responses cached before the rollout (cache TTL ≤ 7 days) — only in those two cases is a total obtained by summing complete (untruncated) rows. Null groups arrive explicitly labeled [null] and are real groups; read null counts from that bucket. Defensive filter: keep only rows in which every requested dimension key is present — a roll-up row omits one or more keys entirely, whereas a genuine null is present with the value [null]. On a correct response this is a no-op; it guards against a stale cached response still carrying legacy subtotal and grand-total rows, each of which equals the whole total and would inflate any sum. When COUNT-ing rows per group, aggregate a different field than the GROUP BY dimension itself — a same-field COUNT of the grouped dimension can 500.
5. Truncated results. Any data tool may return {"data": [...], "truncated": true, "total_rows": N, "returned_rows": M, "guidance": "..."} when the result exceeds the response size limit (~50 KB). The data prefix is incomplete — never compute totals, shares, or trends from it, and never present it as the full result. On aggregations the top-level totals field is unaffected by truncation (computed across all groups, not just the returned prefix) — read grand totals from it instead of re-fetching. Narrow the query (fewer dimensions, more filters, fewer selected columns — or a business metric for a named KPI) and re-fetch only when the rows themselves are needed beyond the cap; with totals present, a SUM/COUNT/MIN/MAX grand total never requires a re-fetch or chunking by dimension (AVG, COUNT_UNIQUE and UNIQUE_VALUES never read totals — true average = SUM total ÷ COUNT total from two calls; true distinct count = the distinct-values tools). A truncated response without totals (pre-rollout cache) cannot answer a grand-total question from its prefix. Re-run the aggregation once — a fresh run may miss the stale entry and return totals. If the re-run still carries no totals, stop re-running and fall back to narrowing or chunking by dimension until the responses are complete, then sum those rows. Never total the prefix.

In particular, never test a control against a budget-named scenario filter without first confirming it exists in the discovered scenario domain — if the plan side lives in a planning-version-like field, route budget-related evidence through that field's versions instead, and record the actual scenario/version used in the evidence package and audit trail.

Datarails Brand Styling

When generating Excel or PowerPoint files, apply Datarails brand styling:

Font: Poppins (fall back to Calibri if unavailable). Weights: 400 regular, 600 semibold, 700 bold.

Colors:

Excel layout:

• Content starts at column B (column A is a narrow gutter)
• Rows 1-6: header banner with navy background, white title text, white subtitle
• Gridlines OFF. Freeze panes at B7.
• Footer as last row with generation date
• Every cell must have font, fill, alignment, and number format set

Number formats: _(* #,##0_);_(* (#,##0);_(* "-"_);_(@_) (default), $#,##0 (dollars), $#,##0.0,,"M" (millions), 0.0% (percent)

Variance coloring: Any cell showing a delta/change: green (2ECC71) if favorable, red (E74C3C) if unfavorable. Apply automatically based on value sign and metric context.

PowerPoint: Navy (0C142B) background, 16:9 widescreen, Poppins font, white text, amber (FFA30F) accent lines, card backgrounds 001F37.

DR.GET Formulas — Authoring Contract

If asked to add live / refreshable Datarails formulas (DR.GET) to a generated workbook, the only valid form is:

=DR.GET(Value, "[DimensionName]", CellRef, "[DimensionName]", CellRef, ...)

• Never transliterate an MCP/API call into a formula. DR.GET takes no table, field, or aggregation arguments — =DR.GET(Value,"financials","Amount","SUM",...) is invented syntax that the Datarails Add-in cannot parse or refresh.
• Dimension names go in square brackets inside quotes ("[Scenario]"). Dimension values are always cell references, never hardcoded strings.
• Date cells referenced by formulas hold end-of-month date serials computed from the calendar — never raw epoch timestamps from API responses (epochs land a day early with a time component and never match).
• Before writing any formula, create the workbook-scoped defined name Value referring to the string constant "Value" (wb.defined_names.add(DefinedName("Value", attr_text='"Value"'))) — otherwise Excel autocorrects the bare token to its built-in VALUE() and the formula breaks.
• Bare =DR.GET(...) only — never wrapped in IFERROR/IF/ROUND.
• Every rule here applies to the retrieval/period family — DR.GET, DR.QTD, DR.YTD, DR.MTD share one form (=DR.QTD(Value, "[Dim]", CellRef, ...)), one cell-reference discipline, one Value defined-name requirement, one no-wrapping rule. "DR.GET" in this contract means that family. Helper functions with their own documented signatures (e.g. DR.INCLUDE, DR.RANGE) are not covered here — author those only from their own documentation, never by analogy with this form.
• In a live Excel context, writing DR formulas and refreshing them is one atomic step — a freshly written DR cell reads Missing until an agent refresh lands, and only read-back values may be quoted. The Excel-context routing preamble (or the skill's own Step 0 workflow) owns that procedure; this contract owns the formula text.

The get-formula skill (/dr-get-formula) is the full reference — parameter cells, validated dimension values, report layouts. Prefer it for whole formula workbooks; apply this contract when adding any retrieval/period DR formula (DR.GET/DR.QTD/DR.YTD/DR.MTD) to a workbook here.

Output

PDF Report

• Scope statement — in-scope check families and the out-of-scope disclaimer (access control / change management / ITGC), on page one
• Executive summary
• Check results (pass/fail/skipped per check, labeled with period + scenario)
• Exception findings
• Recommendations
• Management response section
• Check descriptions with the queries behind each result (appendix)

Excel Evidence Workbook

Sheets map one-to-one to the in-scope checks:

• Summary — check status (pass/fail/skipped) and the scope statement (period + scenario/version actually used)
• Completeness — period coverage vs expected periods; scenario/period checksum results
• Reconciliation — the /dr-reconcile four-check results (including any noted skips)
• Mapping Integrity — parent/child roll-up detail with [null]-bucket (unmapped account) flags
• Samples — row-level detail behind material buckets, traceable to source line items
• Exceptions — findings with severity and recommended action (if any)
• Out of scope — requires external evidence — mandatory sheet listing the control families this tool cannot test (access control, change management, IT general controls) and where their evidence must come from (system administration / IT), so no reader mistakes this package for full SOX coverage

Examples

Q4 2025 year-end evidence package

/dr-audit --year 2025 --quarter Q4

Mid-year evidence package

/dr-audit --year 2025 --quarter Q2

Custom output locations

/dr-audit --year 2025 --quarter Q4 \
  --output-pdf audits/audit_q4_2025.pdf \
  --output-xlsx audits/evidence_q4_2025.xlsx

Use Cases

Data-side evidence for a SOX 404 program

# Year-end evidence package — in-scope control families only
/dr-audit --year 2025 --quarter Q4

Quarterly data-integrity review

# Regular check of completeness, reconciliation, and mapping controls
/dr-audit --year 2025 --quarter Q3

Management certification support

# Data evidence feeding management's SOX 404 process —
# does NOT substitute for access-control or ITGC testing
/dr-audit --year 2025 --quarter Q4

External Auditor Support

# Provide auditors with the report and traceable evidence
/dr-audit --year 2025 --quarter Q4

Performance

• Execution: ~1-2 minutes
• Data-evidencable checks only — every result backed by a query
• Professional report generation
• Evidence package with explicit scope boundaries

Control Framework

Mapped to COSO components only where FinanceOS data provides the evidence:

• Control Activities (in scope): reconciliation, roll-up validation, checksum integrity
• Risk Assessment (in scope, data-side): completeness and mapping integrity of the reported figures
• Information & Communication (in scope): documented evidence trail — the query behind every result
• Monitoring (partial): recurring runs of this package over each close
• Control Environment — access controls, segregation of duties, change management (out of scope): requires system-administration evidence this tool cannot reach

Report Contents

Executive Summary

• Scope statement: in-scope check families + out-of-scope disclaimer
• Audited period and scenario/version actually used
• Check result summary
• Key findings

Detailed Findings

• Exceptions identified
• Severity assessment
• Management response area

Check Descriptions

• Each check's objective
• Query performed (tool + parameters, so the result is reproducible)
• Evidence collected
• Status conclusion (pass/fail/skipped — skips are stated, never faked)

Management Response

• Area for management to address findings
• Remediation timeline
• Owner assignment

Evidence Package

Check Summary Sheet

• Check ID and name
• Objective tested
• Result (pass/fail/skipped)
• Evidence gathered (the query behind the result)

Exception Log

• Finding description
• Severity level
• Supporting evidence
• Recommended action

Supporting Schedules

• Transaction samples (row-level detail behind material buckets)
• Reconciliation details (per-check results from /dr-reconcile)
• Roll-up and checksum workings

Out of Scope — Requires External Evidence

• Access control, change management, IT general controls
• Why: no audit-log, access-history, or user-activity endpoint on the FinanceOS MCP surface
• Where their evidence must come from: system administration / IT

Recommendations

Professional audit recommendations:

• Continue monthly reconciliation procedures (/dr-reconcile)
• Investigate and remap accounts flagged in the [null] roll-up bucket
• Re-run this evidence package each close to keep evidence current
• Source access-review and change-management evidence from system administration — this package cannot supply it

Integration

Works with:

• /dr-anomalies-report - Data quality validation
• /dr-reconcile - Consistency checking
• /dr-extract - Data extraction
• /dr-dashboard - Control monitoring

Related Skills

• /dr-reconcile - Ongoing reconciliation
• /dr-anomalies-report - Data quality
• /dr-extract - Data sourcing
• /dr-dashboard - KPI monitoring