⚠️ This column is a sharing of the prompt creation process for free educational purposes, and is not any investment, legal, or financial advice. AI execution results may contain errors and independent verification is essential, and all responsibility for use is entirely attributed to the user. It cannot be used directly for business decision-making. [View all mandatory confirmation items below the 30-week timeline before use[]](https://www.notion.so/30-29f84ddcec3c80a1a87eda9883d81453?pvs=21)
In Week 11, we established the STEP 0 concept. We learned that Product_Article.md must be read first to enable 100% mapping. Product was the standard, and Economic was a structure that followed Product.
Now we had to create a concrete checklist. At first, we thought we could just take the Product_Checklist.md structure as is. We thought borrowing only PART A and B would be sufficient.
First, we needed to clarify which statistical sources to find economic data from and what search strategy to use for each source. Product extracted product lists from government agency documents, but Economic had to collect economic data from statistical agencies, so we started by organizing the sources.
While writing PART A, we also had to design source-specific strategies for PART B. "Which statistical sources should we find economic data from?" We reviewed with Sequential Thinking. 'How many sources provide economic data? What strategy is needed for each source?’
Claude analyzed. "We need 5 sources that provide trade statistics, import data, GDP contribution, market size, and employment data. We need to use different search strategies for each source.”
The first source was an agency that provides trade statistics. It provides export-import data by item. The search strategy was to search by product name. We search one by one with product names secured from Product_Article.md. We expected about 15-20 searches.
The second source was an agency that provides import data. It provides import amounts for items subject to tariffs. The search strategy was to search by HTS code. However, since we don't have HTS codes yet, we decided to search by product name for now. We expected about 15-20 times.