下载地址
https://www.zhengdianxue.com/wp-content/uploads/skills/arc-zend--amazon-fba-replenishment.zip将提示词发送给你的 AI 安装该 skills
请根据 https://skillhub.cn/install/skillhub.md，安装 @arc-zend/amazon-fba-replenishment。按多 SKU 销量、当前可售库存和分批可售日期模拟逐日库存，识别到货前断货风险，比较需求增长与到货延误情景。支持有货日销量均值、指定日期需求、MOQ 与箱规取整、单 SKU 预算和采购数量上限，导出 Excel 补货清单与库存时间轴。
阅读全文 （展开查看完整说明）
概述
Amazon FBA Replenishment
Turn inventory and lead-time inputs into an explainable reorder decision.
Workflow
- Confirm the SKU, as-of date, demand period, inventory states, production and transport lead time, safety-stock policy, target cover, MOQ, and pack size.
- Read [references/input-fields.md](references/input-fields.md). Keep fulfillable, inbound, reserved, researching, and unfulfillable inventory distinct.
- Do not assign recoverability to reserved or researching units unless the user provides a rate.
- Run
scripts/replenishment.pywith a JSON input. Treat warnings as unresolved planning risks, not footnotes. - Compare at least a base case with any user-requested demand or lead-time scenario. Never manufacture a seasonality multiplier.
- Explain the decision in operational terms and identify the input that most changes the result.
- Prefer the dated batch workflow in [references/batch-planning.md](references/batch-planning.md) for operational decisions. Distinguish arrival at port/warehouse from availability for sale. Only confirmed, dated receipt units are credited on their availability date; do not also add their inbound aggregate.
- Use –out NEW_DIRECTORY with batch JSON for replenishment.xlsx, plan.json and summary.md. Rank urgent pre-arrival shortages separately from routine purchases. Show unconstrained quantity, feasible quantity, residual shortage and the user's limiting assumption.
Batch mode accepts documented daily demand or dated in-stock sales observations, date-specific overrides and explicitly named alternative scenarios. It does not auto-invent promotions or seasonality. Review-period/cover definitions and start-of-day receiving assumptions are in the batch reference. For sparse history or new products, label the demand assumption provisional.
Python 3.10+; JSON planning uses the standard library. Excel export needs scripts/requirements.txt. The host reads user reports and maps them into the input contract; the script does not download Amazon reports or directly parse seller spreadsheets. The existing single-SKU interface is retained as an aggregate estimate, not a dated forecast.
Deliverable
Provide inventory position, adjusted daily demand, days of cover, safety stock, reorder point, latest order date, projected stockout date, raw requirement, MOQ/pack-rounded order quantity, warnings, and scenario comparison.
If critical inputs such as sales velocity or lead time are absent, return a data request and a provisional structure instead of a confident replenishment order.
Boundaries
The calculation is a planning estimate. It does not account for unknown transfer delays, stranded inventory, future promotions, suppressed listings, attribution errors, or Amazon receiving variability. Do not place orders or create shipments.
User-supplied max_order_units is a purchase cap, not a verified Amazon storage-capacity limit. Costs are per-unit procurement costs in an explicitly named currency, not landed profit. Per-SKU budgets are independent; no portfolio budget allocation or cross-currency total is implied. Scenarios are alternative plans and must not be added together. Verify any current Amazon capacity, fees or policy requirements separately from official sources when the user asks.
