Improve codebase architecture

Source: mattpocock/skills — engineering/improve-codebase-architecture

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.

Glossary

why-not-mechanizable: skill workflow guidance; each step requires understanding the surrounding context (repo, task shape, prior state).

Use these terms exactly in every suggestion. Consistent language is the point — don't drift into "component," "service," "API," or "boundary." Full definitions in LANGUAGE.md.

• Module — anything with an interface and an implementation (function, class, package, slice). (review-time: see section note)
• Interface — everything a caller must know to use the module: types, invariants, error modes, ordering, config. Not just the type signature. (review-time: see section note)
• Implementation — the code inside. (review-time: see section note)
• Depth — leverage at the interface: a lot of behaviour behind a small interface. Deep = high leverage. Shallow = interface nearly as complex as the implementation. (review-time: see section note)
• Seam — where an interface lives; a place behaviour can be altered without editing in place. (Use this, not "boundary.") (review-time: see section note)
• Adapter — a concrete thing satisfying an interface at a seam. (review-time: see section note)
• Leverage — what callers get from depth. (review-time: see section note)
• Locality — what maintainers get from depth: change, bugs, knowledge concentrated in one place. (review-time: see section note)

Key principles (see LANGUAGE.md for the full list):

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

This skill is informed by the project's domain model. The domain language gives names to good seams; ADRs record decisions the skill should not re-litigate.

Process

1. Explore

Read the project's domain glossary and any ADRs in the area you're touching first.

Then use the Agent tool with subagent_type=Explore to walk the codebase. Don't follow rigid heuristics — explore organically and note where you experience friction:

• Where does understanding one concept require bouncing between many small modules? (review-time: see section note)
• Where are modules shallow — interface nearly as complex as the implementation? (review-time: see section note)
• Where have pure functions been extracted just for testability, but the real bugs hide in how they're called (no locality)? (review-time: see section note)
• Where do tightly-coupled modules leak across their seams? (review-time: see section note)
• Which parts of the codebase are untested, or hard to test through their current interface? (review-time: see section note)

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

2. Present candidates

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

• Files — which files/modules are involved (review-time: see section note)
• Problem — why the current architecture is causing friction (review-time: see section note)
• Solution — plain English description of what would change (review-time: see section note)
• Benefits — explained in terms of locality and leverage, and also in how tests would improve (review-time: see section note)

Use CONTEXT.md vocabulary for the domain, and LANGUAGE.md vocabulary for the architecture. If CONTEXT.md defines "Order," talk about "the Order intake module" — not "the FooBarHandler," and not "the Order service."

ADR conflicts: if a candidate contradicts an existing ADR, only surface it when the friction is real enough to warrant revisiting the ADR. Mark it clearly (e.g. "contradicts ADR-0007 — but worth reopening because…"). Don't list every theoretical refactor an ADR forbids.

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, drop into a grilling conversation. Walk the design tree with them — constraints, dependencies, the shape of the deepened module, what sits behind the seam, what tests survive.

Side effects happen inline as decisions crystallize:

• Naming a deepened module after a concept not in CONTEXT.md? Add the term to CONTEXT.md — same discipline as /grill-with-docs (see CONTEXT-FORMAT.md). Create the file lazily if it doesn't exist. (review-time: see section note)
• Sharpening a fuzzy term during the conversation? Update CONTEXT.md right there. (review-time: see section note)
• 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 don't re-suggest it?" Only offer when the reason would actually be needed by a future explorer to avoid re-suggesting the same thing — skip ephemeral reasons ("not worth it right now") and self-evident ones. See ADR-FORMAT.md. (review-time: see section note)
• Want to explore alternative interfaces for the deepened module? See INTERFACE-DESIGN.md. (review-time: see section note)