Claude Code checks five-item reorder calculations
ifjuku reports an executed five-item inventory-planning test that identified three order-today items, saved calculation results and corrected rounding that could trigger premature orders.
note.com
Original prompt
Shared by the source author. See source notes for context.
このフォルダの inventory.json を使って、在庫の残り日数と発注日が分かる管理ツールを作ってください。
Reach people exploring AI agents
See placements →What happened
ifjuku reports an executed five-item inventory-planning test that identified three order-today items, saved calculation results and corrected rounding that could trigger premature orders.
Reported by ifjukuHow the agent did it
- Load the five-item sample inventory file in Claude Code.
- Specify usage, lead time and safety stock for each item.
- Calculate reorder thresholds and save a separate results file.
Where a person stepped in
The author checks the formulas and recommends human ordering after manual sample verification. The executed dataset is a demonstration, not actual procurement.
ifjuku asks Claude Code to turn a five-item inventory JSON file into a reorder-planning view. The executed sample contains copy paper, packing tape, black printer ink, envelopes and label stickers; it is a demonstration rather than evidence of real supplier orders.
The instruction specifies current stock, average daily usage, delivery lead time and safety stock. The reorder point is usage multiplied by lead time, plus reserve stock. Items at or below that point are due today; others receive the future date when stock reaches it. The agent saves results to a separate JSON file as well as building the view.
The run flags three items for today and schedules the other two six and seven days later. A calculation pitfall appears during inspection: truncating the time to the reorder point can incorrectly label tomorrow’s item as due today. The correction rounds that interval upward and reserves the today label for the actual threshold condition.
The author recommends checking sample calculations manually before real use and keeping purchasing with a person. Seasonal usage and variable supplier lead times remain inputs to revisit, not facts the agent can infer from quantity alone. The useful result is a checked order-candidate list with explicit formulas and a corrected boundary condition, not autonomous procurement.
Prompt
Original prompt
Source: Blog · ifjuku
Frequently asked questions
Why is stock count insufficient?
The reorder decision also depends on daily usage, delivery lead time and reserve stock.
What rounding error appeared?
Truncating the interval could mark a tomorrow order as due today; the author rounded upward and checked the threshold directly.
Did the agent place orders?
No. The demonstrated result is an order-candidate list, and the author recommends human confirmation and purchasing.
More details
Remaining steps
- Check today-versus-future labels against the source numbers.
- Correct premature rounding and review candidates before any real purchase.
About this account
Claude Code reads inventory data, builds and executes the calculation tool, creates a list view and saved results, and corrects an inspected rounding problem.
Firsthand account by ifjuku; original publication date verified against Primary Japanese byline2026年9月18日07:35; metadata datePublished2026-09-18T07:35:36+09:00. Outcomes are the author’s reports, not independently reproduced. Original Japanese instruction excerpt verified in the public original HTML; surrounding instructions specify formulas, reference date and saved results. English article is AI translated.
Credit: ifjuku
Source published 18 Sept 2026
Source checked 14 Oct 2026
Tools and topics
Claude Code · inventory.json · results.json · Reorder formulas
Reorder boundary checking · Five-item inventory test · Premature-order correction
Editorial completeness: 100/100. Measures documented fields, not agent performance or verified outcomes.