Improve Codebase Architecture

Surface architectural friction and propose deepening opportunities: refactors that turn shallow modules into deep ones.

The aim is testability and AI-navigability. This skill is adapted from Matt Pocock's improve-codebase-architecture skill and aligned to Specorator's glossary, ADR, and stage-gate conventions.

Glossary

Use these architecture terms exactly in every suggestion. Consistent language is the point. Full definitions live in LANGUAGE.md.

• Module: anything with an interface and an implementation.
• Interface: everything a caller must know to use the module: types, invariants, error modes, ordering, configuration.
• Implementation: the code inside.
• Depth: leverage at the interface. Deep means high leverage; shallow means the interface is nearly as complex as the implementation.
• Seam: where an interface lives; a place behavior can be altered without editing in place. Use this term instead of "boundary" unless referring to a documented bounded context.
• Adapter: a concrete thing satisfying an interface at a seam.
• Leverage: what callers get from depth.
• Locality: what maintainers get from depth: change, bugs, knowledge, and verification concentrated in one place.

Key principles:

• Deletion test: imagine deleting the module. If complexity vanishes, it was a pass-through. If complexity reappears across callers, it was earning its keep.
• The interface is the test surface.
• One adapter = hypothetical seam. Two adapters = real seam.

This skill is informed by the project's domain language. Domain glossary entries give names to good seams; ADRs record decisions the skill should not re-litigate.

Process

1. Explore

Read the project's domain language and architecture decisions first:

• docs/glossary/ for domain terms.
• docs/CONTEXT.md or docs/CONTEXT-MAP.md if present.
• docs/adr/ for constraints and decisions.
• The relevant specs/<feature>/design.md or spec.md when the review is tied to active feature work.

Then walk the codebase with the exploration tools available in the current agent environment. In Claude Code, dispatch an Explore subagent when available. In Codex, Cursor, Aider, Copilot, or Gemini, load this file as context and inspect the relevant files directly.

Do not follow rigid heuristics. Explore organically and note where you experience friction:

• Where does understanding one concept require bouncing between many small modules?
• Where are modules shallow: interface nearly as complex as the implementation?
• Where have pure functions been extracted just for testability, but the real bugs hide in how they are called?
• Where do tightly-coupled modules leak across their seams?
• Which parts of the codebase are untested, or hard to test through their current interface?

Apply the deletion test to anything that looks shallow: would deleting it concentrate complexity, or just move it? A "yes, concentrates" answer is the signal.

2. Present candidates

Present a numbered list of deepening opportunities. For each candidate, include:

• Files: which files or modules are involved.
• Problem: why the current architecture is causing friction.
• Solution: plain-English description of what would change.
• Benefits: explain locality, leverage, and how tests would improve.

Use docs/glossary/ and docs/CONTEXT.md vocabulary for the domain, and LANGUAGE.md vocabulary for the architecture. If the glossary defines "Order," talk about "the Order intake module," not a framework-shaped name.

If a candidate contradicts an existing ADR, surface it only when the friction is real enough to warrant revisiting the ADR. Mark it clearly, for example: "Contradicts ADR-0007, but worth reopening because..."

Do not propose interfaces yet. Ask the user: "Which of these would you like to explore?"

3. Grilling loop

Once the user picks a candidate, walk the design tree with them: constraints, dependencies, the shape of the deepened module, what sits behind the seam, and what tests survive.

Side effects happen inline as decisions crystallize:

• Naming a deepened module after a concept not in docs/glossary/? Add the term through the new-glossary-entry skill, or create/update docs/CONTEXT.md if the project uses a context file.
• Sharpening a fuzzy term during the conversation? Update the appropriate glossary or context entry in the same branch.
• User rejects the candidate with a load-bearing reason? Offer an ADR, framed as: "Want me to record this as an ADR so future architecture reviews do not re-suggest it?" Only offer when the reason would help a future explorer avoid the same recommendation. Use the new-adr skill or the repository's ADR template.
• Want to explore alternative interfaces for the deepened module? Use INTERFACE-DESIGN.md.

Do not

• Do not bypass Specorator stage ownership. If this uncovers a missing requirement or architecture decision, update the upstream artifact before implementation.
• Do not propose seams for their own sake. One production adapter without a test or alternate adapter is usually indirection, not depth.
• Do not relitigate accepted ADRs unless the observed friction is strong enough to justify reopening the decision.
• Do not bundle the refactor into unrelated feature work. Turn the selected opportunity into its own task, issue, or PR.