The Agent Skills Directory

[Data Access and Analysis]: The skill reads from local data files such as products.csv, stock_levels.csv, and sales_history.csv located in a designated data directory. This access is necessary for performing demand forecasting and inventory checks.
[Script Execution]: The forecasting logic is implemented in local Python scripts (rolling_mean.py and batch_days_of_cover.py) that are included as part of the skill package. These scripts use standard Python libraries to perform calculations and do not involve any external network operations or third-party dependencies.
[Task Delegation]: The skill provides structured guidelines for when to perform calculations locally versus when to delegate to a specialized subagent. This approach is used to manage context length and ensure analytical accuracy without introducing security risks.

forecasting