domain-model: Keep AI Plans Aligned With Your Codebase Language

Install this skill:

npx skills add mattpocock/skills skill=domain-model -y -g

Source: mattpocock/skills/domain-model

What it does

domain-model stress-tests a plan against the existing domain model of a codebase.

It looks for existing documentation:

• CONTEXT.md, the glossary for a bounded context
• CONTEXT-MAP.md, when a repo has multiple bounded contexts
• docs/adr/, where architectural decisions are recorded

Then it interviews you one question at a time, while also checking the codebase when the answer can be discovered directly.

The important behavior

The skill challenges fuzzy or overloaded language immediately.

If you say "account," it may ask whether you mean Customer, User, Organization, Workspace, or another domain term. If your glossary defines "cancellation" one way and your plan uses it another way, it stops and resolves the conflict.

When a term is resolved, it updates CONTEXT.md inline. It does not wait until the end of the session.

When it writes ADRs

domain-model only offers an ADR when all three are true:

1. the decision is hard to reverse
2. the choice would be surprising without context
3. the choice came from a real trade-off

This keeps ADRs useful instead of turning them into meeting notes.

How it fits the workflow

grill-me → domain-model → to-prd → to-issues → tdd

Use domain-model after the plan has been pressure-tested and before turning it into a PRD.

Pairs well with

• grill-me, to resolve the plan before domain modeling
• to-prd, to synthesize the resolved context into a PRD
• improve-codebase-architecture, which also uses CONTEXT.md and ADRs