Goal
Keep language coherent across code, docs, and multiple repositories.
FPF treats naming as a modeling responsibility: if terms drift, reasoning collapses.
Outputs
- •Update (or create) the main glossary:
- •
.fpf/glossary/glossary.md
- •When mapping terms across contexts (repo A <-> repo B), create a bridge card:
- •
.fpf/glossary/bridges/BRIDGE-${CLAUDE_SESSION_ID}--<slug>.md
Use the template:
- •
.fpf/templates/bridge-card.md
Procedure
- •Detect naming problems Examples:
- •Same word, different meaning across repos.
- •Different words, same meaning.
- •Overloaded words ("process", "service", "model", "spec").
- •Inconsistent abbreviations or casing.
- •Choose a canonical term
- •Prefer stable, descriptive names.
- •Keep aliases for legacy names.
- •Write definitions in context For each term include:
- •definition (in this bounded context),
- •examples (where it appears in code/docs),
- •synonyms/aliases,
- •non-examples (common confusions).
- •Bridge across contexts If two repos have different vocabularies:
- •state source + target contexts,
- •map terms,
- •note loss/distortion (what does not carry over),
- •propose a migration naming plan if needed.
Quality bar
- •Definitions must be locally meaningful (bounded context).
- •Bridges must be explicit; never assume equivalence without a mapping.
- •Update docs/code to reflect canonical naming when safe.