skills/ajrlewis/ai-skills/addon-domain-semantic-adaptation

addon-domain-semantic-adaptation

SKILL.md

Add-on: Domain Semantic Adaptation

Use this skill when generic scaffold outputs should reflect the user's domain language (terminology, labels, workflows, and intent phrasing) without changing core architecture patterns.

Compatibility

  • Works with all architect-*, addon-*, and ui-* skills.
  • Recommended default in production-default mode.

Inputs

Collect:

  • DOMAIN_BRIEF: one-paragraph business/domain context.
  • CANONICAL_TO_DOMAIN_MAP: key term map (for example document->clause packet, user->counsel).
  • PREFERRED_TONE: neutral | formal | technical | friendly.
  • FORBIDDEN_TERMS: terms to avoid in UI/docs.
  • PUBLIC_COPY_SCOPE: ui-only | ui+docs | ui+docs+api-names.

Integration Workflow

  1. Add semantic contract artifacts:
docs/DOMAIN_LANGUAGE.md
docs/DOMAIN_DECISIONS.md
  • DOMAIN_LANGUAGE.md should contain:
  • domain brief
  • canonical-to-domain table
  • forbidden terms
  • naming rules
  1. Apply semantic mapping in generated outputs:
  • UI labels, helper text, empty/loading/error states
  • README and operational docs
  • endpoint and command aliases (when PUBLIC_COPY_SCOPE allows it)
  1. Preserve technical stability:
  • Keep internal architecture and safety controls unchanged.
  • Prefer additive aliases over breaking renames for API/CLI contracts.
  1. Add validation notes:
  • Record mapped terms and intentional non-mapped technical terms in DOMAIN_DECISIONS.md.

Required Template

docs/DOMAIN_LANGUAGE.md

# Domain Language Contract

## Domain Brief
<short context>

## Canonical To Domain Terms
| Canonical | Domain |
|---|---|
| user | counsel |
| document | clause packet |

## Forbidden Terms
- foo
- bar

## Naming Rules
- UI and docs must use domain terms.
- Internal safety/security identifiers may remain canonical.

Guardrails

  • Documentation contract for generated code:

    • Python: write module docstrings and docstrings for public classes, methods, and functions.
    • Next.js/TypeScript: write JSDoc for exported components, hooks, utilities, and route handlers.
    • Add concise rationale comments only for non-obvious logic, invariants, or safety constraints.
    • Apply this contract even when using template snippets below; expand templates as needed.
  • Do not change security-critical semantics for stylistic reasons.

  • Do not rewrite internal identifiers when that would break compatibility.

  • Keep naming consistent across UI, docs, and prompts.

  • Surface ambiguous term mappings as explicit review items.

Validation Checklist

  • Confirm generated code includes required docstrings/JSDoc and rationale comments for non-obvious logic.
test -f docs/DOMAIN_LANGUAGE.md
test -f docs/DOMAIN_DECISIONS.md
rg -n "Canonical To Domain Terms|Forbidden Terms|Naming Rules" docs/DOMAIN_LANGUAGE.md

Manual checks:

  • Generated UI/docs reflect domain terms from user prompt.
  • No forbidden terms appear in user-facing copy.

Decision Justification Rule

  • Every non-trivial decision must include a concrete justification.
  • Capture the alternatives considered and why they were rejected.
  • State tradeoffs and residual risks for the chosen option.
  • If justification is missing, treat the task as incomplete and surface it as a blocker.
Weekly Installs
8
First Seen
14 days ago
Installed on
openclaw8
gemini-cli8
github-copilot8
codex8
kimi-cli8
cursor8