Kando Data
中ENContact us

Re-measuring Inventory Write-downs at XX Technology

FINFRAUD-CN-0411 · Quantitative check · STAR Market · FY2023

As a member of an audit engagement team, pull and reconcile figures across the full annual report, the company's written explanation and the auditor's verification notes to re-measure six inventory write-down amounts on the proper basis.

AuditQuantitative checkCross-table reconciliationEvidence chain
44 files
362Kcharacters of material
6answer leaves
16agent turns in the longest reference rollout
692Kcumulative prompt tokens in one rollout
01

Highlights

The prompt gives no amounts

What to extract, check and compute is left to the model, which must locate the inventory tables, the write-down roll-forward and the by-product and by-ageing detail in 360,000 characters of material.

Management claims vs. independent evidence

The company's explanation is a management representation; only the procedures and conclusions of the auditor and sponsor count as independent audit evidence.

Every number must be recomputable

Each of the six figures needs at least one cross-row, cross-table or cross-period computation, with source and formula shown; a bare conclusion is treated as unverifiable.

From a 1,106-task dataset

Every task in FinFraud-CN-0917 ships with a full Claude Opus 4.8 reference trajectory, and deduplication is zero on all four criteria.

02

Instruction excerpt

An excerpt of the instruction the agent actually receives, showing the task setup and constraints.

instruction.md · 2,085 chars

You are a member of an accounting firm's engagement team. XX Semiconductor Technology Co., Ltd. (STAR Market) saw both inventory and its inventory write-down allowance rise sharply in FY2023. The company has explained in writing the composition and ageing of each inventory item, how net realizable value was determined and how the write-down was computed by product model; the annual report notes disclose inventory by category and the movement of the write-down allowance. A special report on the measurement and presentation of the write-down is required, and you are responsible for the quantitative part.

The company's own explanation is a management representation, not independent audit evidence; the procedures and conclusions in the auditor's or sponsor's documents are. All data needed is in these files; this prompt repeats no amounts.

The question

Were the FY2023 write-down charge, its effect on profit and the year-end presentation measured and presented on a proper basis? Re-measured on a proper basis, what are the following amounts?

  • The inventory write-down loss charged to profit in 2023;
  • The reduction of the allowance from reversals or write-offs;
  • The difference between that charge and the total of the company's by-product-model write-down table;
  • The total write-down re-computed by ageing band for all inventory other than raw materials, using the company's disclosed NRV ratios per ageing band;
  • The year-end carrying amount of work in progress and goods under processing combined;
  • The year-end carrying amount of finished goods and goods shipped combined.

NOTEEnglish translation of the original Chinese instruction; JSON field names and output structure are omitted.

03

Attachments

Everything visible inside the agent container, with statistics generated from the task package. Original files ship with the full sample package.

FileTypeSizeBytes
MATERIALS_INDEX.md.txtMaterials indexmd379 chars829 B
annual_report_1.txtFY2023 annual report in full (with financial statements and notes)txt241K chars515 KB
auditor_reply_1.txtSponsor and auditor verification notestxt57K chars126 KB
company_reply_1.txtCompany's written explanation (management representation)txt63K chars140 KB
04

Scoring design

6numeric answer leaves

Full criteria text, weights and reference answers ship with the complete data package.

6 numeric answer leaves, plus evidence grounding and derivation structure.

  • Numeric leaves are compared to the reference within a relative-error tolerance
  • The evidence dimension checks that sources are complete and figures match the materials
  • The derivation dimension checks for sources, formulas and intermediate values
  • If the grader errors it writes no result and exits non-zero, so infrastructure failures are never scored as wrong answers
05

Model results

Trajectory statistics from frontier models in real agent harnesses. One turn is one model call; the context curve shows how state accumulates over a long-horizon task.

ModelAgent turnsTool callsCumulative prompt tokensPeak contextTimeResult
Claude Opus 4.8Claude Code · headless, offline1615692K61K12m 01sCompleted
038K75Kturn 1turn 16
Claude Opus 4.8Y: context tokens per turn · X: agent turn

Every task in this dataset ships with a full Claude Opus 4.8 reference trajectory (all tool calls and reasoning); trajectories contain no reference answers or scores.

06

Specification

TASK SPECIFICATION
Task IDFINFRAUD-CN-0411
Task formatHarbor task.toml
Agent time limit30 min
Verifier time limit5 min
Agent container1 vCPU · 1 GB memory
NetworkFully offline (empty allowlist)
Scoring isolationSeparate verifier containerJudge credentials are injected only into the verifier and are invisible to the agent
Instruction length2,085 characters
DatasetFinFraud-CN-0917 · 1,106 tasks
Task typeQuantitative check (553 tasks)
Contact

Want to run this task on your own platform?

Leave your details and we will send the full sample package for this task (instruction, attachments, verifier and reference trajectories) and walk you through volume and customisation for similar tasks.

  • Full task package and verifier
  • More tasks in this subdomain
  • Integration and re-run support

REFINFRAUD-CN-0411 · Re-measuring Inventory Write-downs at XX Technology

Used only to follow up on this request.