← Back to Skills Marketplace
openclawvincent

inventory-anomaly

by Vincent_Openclaw · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
542
Downloads
0
Stars
1
Active Installs
1
Versions
Install in OpenClaw
/install inventory-anomaly
Description
库存异常检测和需求预测系统生成工具。当用户需要搭建库存管理系统、实现异常检测算法、开发需求预测功能(如ARIMA模型)、创建库存预警系统时使用此skill。特别适用于制造业、零售业、备件管理等场景,需要处理Excel数据、检测库存异常、预测未来需求并生成报告的情况。
README (SKILL.md)

库存异常检测和需求预测系统

这个skill帮助你快速搭建一个完整的库存异常检测和需求预测系统,类似于电梯配件库存备货提醒系统。

系统架构

系统采用模块化设计,包含以下核心模块:

1. 数据管理模块 (data_manager.py)

  • 读取和写入Excel数据
  • 管理配件信息、库存数据、订单历史
  • 提供统一的数据查询接口

2. 库存查询模块 (inventory.py)

  • 查询单个或所有配件库存
  • 判断库存状态(正常/不足/严重缺货)
  • 生成库存预警信息

3. 需求计算模块 (demand_calculator.py)

  • 基于订单历史计算需求统计
  • 计算日均、周均、月均需求
  • 生成补货建议

4. 异常检测模块 (anomaly_detector.py)

  • 检测库存异常(库存不足、严重缺货)
  • 检测需求异常(超过3倍标准差)
  • 评估异常严重程度

5. 需求预测模块 (predictor.py)

  • 使用ARIMA模型预测未来需求
  • 每周滚动预测
  • 检测预测异常(超过3倍标准差)
  • 生成紧急补货建议

6. 数据生成模块 (generate_fake_data.py)

  • 生成包含各种异常情况的测试数据
  • 支持每天运行一次的数据更新机制
  • 生成正常库存、库存不足、库存超量、需求异常增多等场景

7. 主程序 (main.py)

  • 整合所有功能模块
  • 生成综合报告(TXT格式)
  • 实现每日自动更新机制

使用步骤

步骤1:创建项目结构

首先创建标准的项目目录结构:

your-project/
├── src/                    # 源代码目录
├── test/                   # 测试目录
├── data/                   # 数据目录
├── output/                  # 输出目录
└── spec/                    # 规格说明目录
    ├── ME2AI/             # 人类维护的规格
    └── AI2AI/             # AI维护的代码状态

步骤2:生成核心模块

根据你的业务需求,生成以下模块:

  1. data_manager.py - 数据管理核心
  2. inventory.py - 库存查询功能
  3. demand_calculator.py - 需求计算功能
  4. anomaly_detector.py - 异常检测功能
  5. predictor.py - 需求预测功能(ARIMA模型)
  6. generate_fake_data.py - 测试数据生成
  7. main.py - 主程序入口

步骤3:配置业务参数

根据你的具体业务,调整以下参数:

  • 配件信息:配件ID、名称、规格、单价
  • 库存参数:当前库存、安全库存、最大库存
  • 时间范围:订单历史天数(默认90天)
  • 预测周期:预测未来周数(默认4周)
  • 异常阈值:标准差倍数(默认3倍)
  • 更新频率:每天运行一次

步骤4:生成测试数据

使用generate_fake_data.py生成测试数据:

  • 配件数量:20个(可调整)
  • 订单历史:680条(最近90天)
  • 异常分布:
    • 严重缺货(库存=0):3个配件
    • 库存不足:3个配件
    • 库存超量:4个配件
    • 正常库存:10个配件
    • 需求异常增多:3个配件

步骤5:运行系统

执行主程序:

python src/main.py

系统会:

  1. 检查是否需要更新数据
  2. 如果需要,生成新的测试数据
  3. 执行库存查询、异常检测、需求预测
  4. 生成TXT格式报告
  5. 保存报告到output/report.txt

报告格式

生成的报告包含以下部分:

1. 库存概览

  • 总配件数
  • 异常库存数量

2. 异常检测

  • 库存异常数量
  • 每个异常配件的详细信息:
    • 配件ID和名称
    • 异常类型
    • 现有库存、需求数量、缺货数量

3. 需求预测(未来4周)

  • 成功预测的配件数量
  • 异常提醒的配件数量
  • 每个异常配件的详细信息:
    • 配件ID
    • 预测周次
    • 异常类型和Z分数

4. 紧急补货建议

  • 需要紧急补货的配件数量
  • 每个配件的缺口数量

技术栈

  • 数据处理:pandas
  • 数值计算:numpy
  • 时间序列预测:statsmodels (ARIMA模型)
  • Excel操作:openpyxl

依赖安装

pip install pandas numpy statsmodels openpyxl

扩展建议

自定义配件类型

修改generate_fake_data.py中的配件信息:

parts_info = pd.DataFrame({
    '配件ID': [f'P{i:03d}' for i in range(1, 21)],
    '配件名称': ['你的配件名称1', '你的配件名称2', ...],
    '规格': ['规格1', '规格2', ...],
    '单价': np.random.randint(50, 500, 20)
})

调整异常检测阈值

修改anomaly_detector.py中的标准差倍数:

threshold = 3 * std_demand  # 改为其他值

修改预测模型

predictor.py支持多种预测模型,可以替换ARIMA为:

  • Prophet
  • LSTM
  • XGBoost
  • 其他时间序列模型

自定义报告格式

修改main.py中的generate_txt_report函数,调整报告格式和内容。

常见问题

Q: 如何使用真实的Excel数据?

A: 将你的Excel文件放到data/目录,命名为spare_parts.xlsx,包含三个工作表:

  • 配件信息:配件ID、名称、规格、单价
  • 库存:配件ID、当前库存、安全库存、最大库存
  • 订单历史:订单ID、配件ID、数量、日期

Q: 如何调整预测周期?

A: 修改main.py中的调用参数:

predictions = system.predict_demand(weeks_ahead=8)  # 预测8周

Q: 如何添加新的异常类型?

A: 在anomaly_detector.py中添加新的检测逻辑,例如:

if current_stock > max_stock:
    anomalies.append({
        '配件ID': part_id,
        '异常类型': '库存超量',
        ...
    })

Q: 如何集成到现有系统?

A: 这个系统设计为独立运行,可以通过以下方式集成:

  1. API集成:将各模块封装为REST API
  2. 数据库集成:将data_manager改为使用数据库
  3. 定时任务:使用cron或Windows任务计划器定时运行
  4. 消息通知:添加邮件、短信、钉钉等通知功能

性能优化建议

  1. 数据缓存:频繁读取的数据可以缓存到内存
  2. 批量处理:使用向量化操作替代循环
  3. 并行计算:多配件预测可以并行处理
  4. 增量更新:只处理新增的订单数据

安全注意事项

  1. 数据备份:定期备份Excel数据
  2. 权限控制:限制数据文件的访问权限
  3. 日志记录:记录系统运行日志
  4. 异常处理:完善错误处理和恢复机制
Usage Guidance
This skill appears to do what it claims: local Excel-based inventory anomaly detection and ARIMA forecasting, producing TXT reports. Before installing/running: (1) review and run the code in an isolated environment (virtualenv/container). (2) Ensure your data/spare_parts.xlsx has the required sheets and columns — the code often assumes rows exist and may raise IndexError on missing entries. (3) The main script references helper functions (check_if_run_today, save_run_date) that are not included; expect to add or adjust those bits before production use. (4) Pin dependency versions and run tests with fake data first (templates include a fake-data generator). (5) The package does no network calls or secret access, but it will read/write local files — do not point it at sensitive system directories. If you want, I can list the exact lines/functions that are likely to raise errors or produce exceptions so you can patch them before use.
Capability Analysis
Type: OpenClaw Skill Name: inventory-anomaly Version: 1.0.0 The skill bundle provides a comprehensive template for building a local inventory anomaly detection and demand prediction system using Python, pandas, and ARIMA models. The code consists of project scaffolding scripts (create_structure.py) and modular templates (data_manager.py, predictor.py, etc.) that interact solely with local Excel files for data storage. No evidence of data exfiltration, malicious execution, or harmful prompt injection was found.
Capability Assessment
Purpose & Capability
Name/description match the included templates and scripts: data management, inventory queries, anomaly detection, ARIMA-based forecasting, fake-data generation and TXT report generation. No unrelated binaries or credentials are requested.
Instruction Scope
SKILL.md instructs copying templates, installing Python deps, and running src/main.py; all runtime behavior (reading/writing Excel under data/, generating output/report.txt) is within the stated purpose. Notes: main_template.py references helper functions (check_if_run_today, save_run_date, generate_fake_data invocation location) that are not defined in the templates as provided, and several data-access methods assume records exist which can raise exceptions if input Excel sheets are missing or malformed. These are functional/robustness issues rather than malicious scope creep.
Install Mechanism
The skill is instruction-only (no remote installers). A small requirements.txt lists common Python packages (pandas, numpy, statsmodels, openpyxl) from PyPI — expected for this workload. No downloads from arbitrary URLs or archive extraction are present.
Credentials
No environment variables, credentials, or unusual config paths are requested. All file I/O is local (data/, output/) and aligns with the described purpose.
Persistence & Privilege
always:false and no code attempts to modify other skills or global agent settings. The skill writes project files and report.txt under the project directories only, which is appropriate for a local tool.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install inventory-anomaly
  3. After installation, invoke the skill by name or use /inventory-anomaly
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
Initial release of inventory-anomaly-prediction skill: - Provides a modular system for inventory anomaly detection and demand forecasting, supporting use cases in manufacturing, retail, and spare parts management. - Includes core modules for Excel data management, inventory queries, demand calculation, anomaly detection, ARIMA-based demand forecasting, test data generation, and report creation. - Supports generating and processing real or simulated inventory and order data, identifying abnormal inventory and demand trends, and providing restock recommendations. - Outputs a comprehensive TXT report covering inventory status, detected anomalies, forecast results, and urgent replenishment advice. - Designed for easy customization and integration, with guidance for parameter adjustment, model replacement, and report formatting. - Utilizes pandas, numpy, statsmodels (ARIMA), and openpyxl for data handling and analytics.
Metadata
Slug inventory-anomaly
Version 1.0.0
License MIT-0
All-time Installs 1
Active Installs 1
Total Versions 1
Frequently Asked Questions

What is inventory-anomaly?

库存异常检测和需求预测系统生成工具。当用户需要搭建库存管理系统、实现异常检测算法、开发需求预测功能(如ARIMA模型)、创建库存预警系统时使用此skill。特别适用于制造业、零售业、备件管理等场景,需要处理Excel数据、检测库存异常、预测未来需求并生成报告的情况。 It is an AI Agent Skill for Claude Code / OpenClaw, with 542 downloads so far.

How do I install inventory-anomaly?

Run "/install inventory-anomaly" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is inventory-anomaly free?

Yes, inventory-anomaly is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does inventory-anomaly support?

inventory-anomaly is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created inventory-anomaly?

It is built and maintained by Vincent_Openclaw (@openclawvincent); the current version is v1.0.0.

💬 Comments