本文へ移動
cccskills
無料GitHub で公開

audit-workpaper

Write a clear, reviewable audit workpaper documenting procedures, evidence, and conclusions. Use when an auditor says "write a workpaper", "document this audit procedure", "tick and tie this", "I need to document my testing", "substantive testing workpaper", "controls testing documentation", "audit evidence memo", "prepare the workpaper for this balance", "document the audit steps I performed", or needs to capture any audit work in a format that supports review and sign-off. Also trigger when someone has completed audit testing and needs to write it up even if they don't use the word "workpaper".

インストール方法を見る

含まれるファイル(2)

  • SKILL.md8.2 KB
  • references/assertions-by-area.md7.8 KB

SKILL.md(原文)

インストールする前に、エージェントに与えられる指示の中身を確認できます。

Reference Files

  • references/assertions-by-area.md - Maps each financial statement area (AR, revenue, AP, cash, inventory, fixed assets, debt, equity, accruals) to standard assertions and typical audit procedures. Read this in Step 1 when selecting assertions and in Step 4 when designing procedures for the account being tested.

Overview

Based on "Auditing and Assurance Services" by Arens, Elder, and Beasley - the standard reference for audit methodology. A workpaper is not a record of what you did; it is evidence that you did sufficient, appropriate work to support your conclusion. The reviewer should be able to pick up the workpaper, understand the objective, trace every tick mark back to a source document, and evaluate the conclusion - without asking the preparer a single question.

PCAOB AS 1215 and AICPA AU-C 230 both require that workpapers be complete enough to allow an experienced auditor with no prior connection to the engagement to understand the work performed.

Workflow

Step 1: Write the workpaper header

Every workpaper starts with a standard header that identifies what it covers and where it lives in the audit file.

CLIENT: [client name]
ENGAGEMENT: [audit / review / agreed-upon procedures]
PERIOD END: [date]
WORKPAPER REF: [e.g., B-3, or PBC-12]
AREA: [e.g., Accounts Receivable, Revenue, Cash]
ASSERTION(S) TESTED: [existence / completeness / accuracy / valuation / cutoff / rights & obligations / presentation]
PREPARED BY: [initials]   DATE: [date]
REVIEWED BY: [initials]   DATE: [date]

Assertions are the anchor. Every procedure in the workpaper must link back to the assertion(s) it tests. If you cannot state which assertion a procedure addresses, remove it.

Step 2: State the objective

Write one or two sentences that define what the workpaper is meant to conclude. This drives everything that follows.

Example:

OBJECTIVE
To obtain sufficient appropriate evidence that accounts receivable at [date]
exists, is accurately stated at net realizable value, and represents amounts
owed by customers for goods or services delivered (existence, valuation, rights).

If the objective is vague, the procedures will be unfocused and the conclusion will be unsupportable.

Step 3: Document the source data and population

Before describing procedures, document what you started with.

POPULATION / SOURCE DATA
Source: [e.g., AR aging report exported from [ERP] on [date], provided by [client contact]]
Population total: $[amount]  ([X] items)
Agreed to: [TB line / lead schedule ref]  confirmed
Stratification: [if sampling, document how population was stratified]
Sampling method: [random / haphazard / systematic / full population]
Sample size: [n] - basis: [risk-based / statistical / tolerable misstatement $X]

Tracing the population to the trial balance before testing eliminates a common deficiency: testing from a list that does not agree to the financial statements.

Step 4: Document procedures performed

For each procedure, use this format: procedure - evidence - result.

PROCEDURE 1: [Concise description of what was done]
Evidence: [What document, system, or confirmation was used]
Result: [What was found - agree / exception / N/A]
Tick mark: [assign a symbol, e.g., check = agreed to source, ^ = recalculated]

Example:

PROCEDURE 1: Agreed sample AR balances to signed customer invoices on file.
Evidence: Invoices obtained from client PBC folder [tab: AR Support].
Result: All 25 sample items agreed to invoice amounts. No exceptions.
Tick mark: check = agreed to invoice

PROCEDURE 2: Confirmed 10 balances over $50K directly with customers via
positive confirmation. Sent [date], responses received by [date].
Evidence: Signed confirmation letters filed in [workpaper ref: B-4].
Result: 9 of 10 confirmed without exception. 1 difference noted - see exception below.

EXCEPTION: Customer [X], confirmed balance $142,500 vs. book $148,000. Difference
of $5,500. Client explanation: credit memo issued [date], not yet posted. Credit memo
reviewed and agreed to [doc]. Resolved - no misstatement.

Step 5: Summarize tick mark legend

List every symbol used in the workpaper with its definition.

TICK MARK LEGEND
check  = Agreed to source document (invoice, contract, bank statement)
^      = Independently recalculated
diamond = Confirmed directly with third party
F      = Footed (column totals verified)
CF     = Cross-footed (row and column totals reconciled)
TB     = Agreed to trial balance / lead schedule
PY     = Agreed to prior year workpaper [ref]

Every tick mark in the workpaper body must appear in this legend. Unexplained tick marks are a common review finding.

Step 6: Write the conclusion

The conclusion must directly answer the objective. It is not a summary of what was done.

CONCLUSION
Based on procedures performed, we obtained sufficient appropriate evidence that
accounts receivable of $[amount] at [date] exists, is accurately stated, and
represents valid customer obligations (existence, valuation, rights assertions).
No exceptions were noted that, individually or in aggregate, represent a material
misstatement. This balance is supported for audit sign-off.

If exceptions were found, state whether they were resolved and whether they affect the conclusion. Never write a clean conclusion when there are unresolved exceptions.

Anti-Patterns

1. Procedure without result Bad: "Reviewed accounts receivable aging for unusual items." Good: "Reviewed accounts receivable aging report for items over 90 days and any credit balances. Identified 3 balances over 90 days totaling $28K. Obtained client explanation for each - see detail below. No exceptions noted." Procedures without results are not audit evidence.

2. Conclusion that does not match the objective Bad: Objective tests existence; conclusion says "balances are reasonable." Good: Conclusion directly addresses the assertion(s) in the objective: existence, accuracy, valuation. Mismatched objectives and conclusions signal a copy-paste error or that the conclusion was written before the work.

3. Population not agreed to the trial balance Bad: Testing a sample from the AR aging without confirming the aging total agrees to the GL. Good: Always foot the population and agree it to the trial balance before selecting the sample. A difference between the aging and the GL must be resolved before testing.

4. Unexplained exceptions left in the file Bad: "Difference of $5,500 noted. Not investigated." Good: Every exception must be investigated, explained, and resolved. Either it is a misstatement (quantify it, add to the summary of misstatements), or it is explained and cleared. Unresolved exceptions cannot be signed off.

5. Workpaper requires the preparer to explain it Bad: A workpaper so brief or cryptic that the reviewer has to ask what it means. Good: The reviewer should be able to pick up the workpaper and understand the objective, evidence, and conclusion without asking a question. If that test fails, add more detail.

Quality Checklist

  • Header complete: client, period, ref, assertions, preparer, reviewer
  • Objective states the specific assertion(s) being tested
  • Population agreed to trial balance before testing
  • Source of population documented (system, date, who provided)
  • Each procedure has: description, evidence source, and result
  • All tick marks defined in a legend
  • Exceptions documented with investigation and resolution
  • Conclusion directly answers the objective
  • Conclusion is clean only if all exceptions are resolved
  • Workpaper is self-contained - no reviewer questions required

レビュー

まだレビューはありません。使ってみた感想をお寄せください。

同じリポジトリのスキル

概要と使いどころ

Rate any product, feature, or experience on the 11-star scale (Brian Chesky's Airbnb thought experiment). Use when user says "rate this experience", "11-star", "star rating", "experience audit", "how good is this", "experience rating", "product audit", "quality assessment", or wants to evaluate product quality and identify improvement paths. Also trigger when user wants to benchmark a feature, assess where a product stands, or map out what "great" looks like - even if they don't explicitly say "11-star".

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Generate A/B copy variants grounded in Eugene Schwartz's awareness and market-sophistication stages and John Caples' tested-headline discipline. Every variant declares its angle, the awareness stage it targets, and a falsifiable hypothesis, and the skill recommends which pair to test first and what delta counts as signal. Use when the user says "A/B test this", "write variants", "give me options", "test different angles", "copy variants", "ab copy", "test this headline", "multiple versions", "which version is better", or wants to test copy before publishing. For reviewing copy that already exists, see copy-review. For sample size and stopping rules, see growth-experiment. For diagnosing a page rather than its words, see page-cro. For full paid ad sets by platform, see ad-campaign-writer.

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Write measurable acceptance criteria for requirements or user stories. Use when the user says "write acceptance criteria", "definition of done", "how do we test this", "fit criteria", "given when then", "Gherkin scenarios", "what does done look like", "testable conditions", "how will we know this works" - even if they don't explicitly say "acceptance criteria".

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Build a strategic account plan for winning, retaining, or expanding a named account. Use when the user says "build an account plan", "strategic account planning", "how do I grow this account", "account expansion strategy", "whitespace analysis", "QBR prep", "key account review", "map the stakeholders in this account", "land and expand plan", "how do I break into [company]", or wants a structured strategy for a specific account.

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Write paid ad copy for LinkedIn, Google, Meta, and YouTube. Produces multiple variants per platform, each targeting a different Eugene Schwartz awareness level and Cialdini persuasion principle. Use when the user asks for ad copy, ad variants, "write LinkedIn ads", Google ads, Meta ads, paid social copy, ad headlines, or wants creative for a paid campaign. Reads brand voice, ICP, and positioning from knowledge/. For organic social posts, see linkedin-post or social-calendar. For the landing page the ad points at, see landing-page-writer. For testing variants, see ab-copy-writer.

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

Design clean and consistent APIs. Use when the user says "design an API", "API design review", "design these endpoints", "REST API for X", "GraphQL schema", "API contract", "how should this endpoint work", "review my API design", or wants to create or improve an API interface - even if they don't explicitly say "API design".

日本語の概要は準備中です。原文の説明を表示しています。

qa-aman/claude-skills202026年9月10日 更新

qa-aman のスキルをすべて見る

このスキルの問題を報告する