excel

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Chinese

<modeling_standards>

Zero Formula Errors: Models MUST have zero #REF!, #DIV/0!, or #VALUE! errors.
Dynamic Logic: You MUST NOT hardcode derived values. You MUST use Excel formulas for all calculations.
Assumptions: You MUST place all inputs in dedicated assumption cells. </modeling_standards>

<professional_formatting>

Standards: Specify units in headers ("Revenue ($mm)"). Format zeros as "-".
Color Coding: The agent SHOULD follow the project's
```
branding
```
skill for color choices. If not defined, the agent SHOULD default to professional standards (e.g., Blue for hardcoded inputs, Black for formulas).
Visuals: You SHOULD use
```
artifact_tool
```
to render sheets and verify layout. Reference:
```
references/artifact_tool_spreadsheets_api.md
```
. </professional_formatting>

<technical_workflows>

<modeling_standards>

<professional_formatting>

格式标准：在表头中指定单位（例如“收入（$mm）”）。将零值格式化为“-”。
颜色编码：Agent应遵循项目的
```
branding
```
skill选择颜色。如果未定义，则应默认采用专业标准（例如，蓝色用于硬编码输入，黑色用于公式）。
可视化：应使用
```
artifact_tool
```
渲染表格并验证布局。参考：
```
references/artifact_tool_spreadsheets_api.md
```
。 </professional_formatting>

<technical_workflows>

You SHOULD use Pandas for heavy lifting and aggregation.
You SHOULD convert to Openpyxl for final professional formatting and formula insertion.

Before delivery, you MUST run the audit script:

```
python scripts/recalc.py output.xlsx
```
You MUST fix all errors identified in the resulting JSON summary. </technical_workflows>

<citation_logic>

Citations: You SHOULD cite sources for hardcoded data in cell comments.
Best Practices: See
```
references/spreadsheet.md
```
for guidance on cross-sheet references and complex formula construction. </citation_logic>

</excel_professional_suite> </instructions>

交付前，必须运行审计脚本：

<citation_logic>

引用规则：应在单元格注释中引用硬编码数据的来源。
最佳实践：有关跨表引用和复杂公式构建的指导，请参阅
```
references/spreadsheet.md
```
。 </citation_logic>

</excel_professional_suite> </instructions>