domain-modeling

Actively build and sharpen a project's domain model — challenge fuzzy terms, stress-test with scenarios, and write CONTEXT.md and ADRs inline as decisions crystallise. Use when the user wants to pin d

SKILL.md

GitHub

Domain Modeling

Actively build and sharpen the project's domain model as you design. This is the active discipline — challenging terms, inventing edge-case scenarios, and writing the glossary and the decisions down the moment they crystallise.

Merely reading CONTEXT.md for vocabulary is not this skill; that's a one-line habit any skill can do. Reach for this skill when you are changing the model, not just consuming it.

File structure

Most repos have a single context:

/
├── CONTEXT.md
├── docs/
│   └── adr/
│       ├── 0001-event-sourced-orders.md
│       └── 0002-postgres-for-write-model.md
└── src/

If a CONTEXT-MAP.md exists at the root, the repo has multiple contexts. The map points to where each one lives:

/
├── CONTEXT-MAP.md
├── docs/
│   └── adr/                          ← system-wide decisions
├── src/
│   ├── ordering/
│   │   ├── CONTEXT.md
│   │   └── docs/adr/                 ← context-specific decisions
│   └── billing/
│       ├── CONTEXT.md
│       └── docs/adr/

Create files lazily — only when you have something to write. If no CONTEXT.md exists, create one when the first term is resolved. If no docs/adr/ exists, create it when the first ADR is needed. Full layout rules: CONTEXT-FORMAT.md.

During the session

Challenge against the glossary

When the user uses a term that conflicts with the existing language in CONTEXT.md, call it out immediately. "Your glossary defines 'cancellation' as X, but you seem to mean Y — which is it?"

Sharpen fuzzy language

When the user uses vague or overloaded terms, propose a precise canonical term. "You're saying 'account' — do you mean the Customer or the User? Those are different things."

Discuss concrete scenarios

When domain relationships are being discussed, stress-test them with specific scenarios. Invent scenarios that probe edge cases and force the user to be precise about the boundaries between concepts.

Cross-reference with code

When the user states how something works, check whether the code agrees. If you find a contradiction, surface it: "Your code cancels entire Orders, but you just said partial cancellation is possible — which is right?"

What this checks, and what it doesn't. The code, the committed CONTEXT.md, and the ADRs — nothing else. A naming collision that was argued out and deliberately settled in a closed ticket months ago will be surfaced as if it were new, because the tracker isn't read. When the user says "we settled this," take it as settled and offer to record the resolution rather than re-litigating it.

Update CONTEXT.md inline

When a term is resolved, update CONTEXT.md right there. Don't batch these up — capture them as they happen. Use the format in CONTEXT-FORMAT.md. The batched version is a summary of a session; the inline version is the session's actual output.

CONTEXT.md is a glossary and nothing else. This is the rule that breaks in the field: "write it to CONTEXT.md" reads as permission to persist every answer the user gives, and the file grows into a running spec. Write a term only when it clears both bars:

  1. It names a concept specific to this domain — not a general programming concept (timeout, retry, error type) however heavily the project uses it.
  2. The entry says what the thing is, in one or two sentences, with the rejected synonyms under _Avoid_.

Everything else — schemas, endpoints, decisions, open questions, session notes — belongs somewhere that isn't this file. A healthy CONTEXT.md gets shorter about as often as it gets longer; if it's past a few hundred lines, it has absorbed material that was never glossary content, and the repair is a pass that strips implementation detail rather than a split into more files.

Offer ADRs sparingly

Only offer to create an ADR when all three are true:

  1. Hard to reverse — the cost of changing your mind later is meaningful
  2. Surprising without context — a future reader will wonder "why did they do it this way?"
  3. The result of a real trade-off — there were genuine alternatives and you picked one for specific reasons

If any of the three is missing, skip the ADR. Use the format in ADR-FORMAT.md.

Quality Loop

Before ending a modelling pass, check the artifacts against what the session actually settled:

  1. Every resolved term is written down — nothing settled in conversation is left only in the transcript.
  2. Definitions say what a thing IS — not what it does, not how it's implemented.
  3. Rejected words are recorded — each term carries the _Avoid_ list that keeps the language opinionated.
  4. No general programming concepts — timeouts, retries, and error types are not domain terms.
  5. Each ADR passes all three gates — hard to reverse, surprising without context, a real trade-off.
  6. The file didn't just grow — anything added that isn't a term, and anything already there that stopped being one, comes back out in the same pass.

If two or more fail, fix the artifacts and re-check. Cap at two passes.

Delegation Map

User need Delegate via Notes
Stress-test a plan against the domain model Invoke grill-with-docs That skill runs the interview and calls this discipline inline
Name a deepened module in the domain's language Invoked by codebase-design / codebase-improve-architecture Resolve the term here, then name the module after it
Turn the settled model into a written spec Invoke plan-to-spec The glossary supplies the spec's vocabulary

Loop eligibility: false — the model sharpens when the design conversation happens, not on a schedule.

Host portability: this skill only reads the repo and writes markdown — no host-specific tools. See PORTABILITY.md.

<!-- self-evolve:start -->

Self-Evolve Loop

Journal ~/.ink-and-agency/learnings/domain-modeling.md (or workspace-local .ink-and-agency/ where the sandbox confines writes). Read it first; append what the run taught — SELF-EVOLVE.md.

<!-- self-evolve:end -->