Skip to content

Workbook reference

The model is usually defined through a spreadsheet workbook. This section details the data input requirements.

Workbook Sheets

Mandatory sheets:

  • Impact Multipliers - imp Define the uncertain impact multipliers, their distributions, and how they combine with characteristics.
  • Correlation Matrix - cov Define correlations between impact multipliers, and optionally between opportunity noise modifiers.
  • Characteristics - fos Define how each opportunity scores on each impact multiplier.
  • Funding Parameters - meta Define the budget, time horizon, and optional placeholder-opportunity settings.
  • Opportunity Funding - orgmeta Define room for funding, diminishing-returns parameters, year availability, and model coverage for each opportunity.

Optional sheets:

  • Formulas - formulas Define derived variables using formulas instead of sampling them directly.

Optional supporting sheets:

  • Custom Distributions - separate sheets referenced from imp, not a literal tab named custom_distributions Provide empirical value lists for multipliers that use Distribution = custom.

Parser Load Order

The engine initialises workbook inputs in this order:

  1. formulas
  2. imp
  3. cov
  4. fos
  5. meta
  6. orgmeta

This matters because later sheets can depend on information or defaults established earlier. For example:

  • cov is filtered against the final multiplier set.
  • orgmeta uses meta$Years to validate Available in year X.
  • fos can reference formula-generated multipliers that are added to the multiplier table later.

How To Use These Pages

Each page in this section is written as a parser contract:

  • accepted structure,
  • defaults and inferred values,
  • validation behaviour,
  • implementation details that are easy to miss when reading only the spreadsheet examples.

If you are updating example workbooks or creating agent prompts, prefer these pages over older prose descriptions.