codebase-design

name	codebase-design
description	Use when designing or restructuring a module and reaching for the shared deep-module vocabulary — interface, seam, depth, adapter, leverage, locality — to decide where a boundary goes, whether an abstraction earns its keep, or how to make a module testable through its interface. Also when another skill needs these terms. Triggers on: "deep module", "where should this seam go", "is this abstraction worth it", "make this testable", "module interface", "reduce coupling".
license	MIT (adapted from mattpocock/skills)

Codebase Design

Design deep modules: a lot of behaviour behind a small interface, placed at a clean seam, testable through that interface. Use this language and these principles wherever code is being designed or restructured. The aim is leverage for callers, locality for maintainers, and testability for everyone.

Before You Build — The Ladder

Before writing any code, climb this ladder. Stop at the first rung that holds:

1. Does this need to exist? If not, delete the requirement, not the code. YAGNI is not a suggestion — it's a discipline.
2. Does stdlib or the platform already do this? Use it. Don't wrap what's free. See Platform Native for a checklist.
3. Does an already-installed dependency do this? Use it. No new pip install or npm add for what you already ship.
4. Can it be one line? One line. No abstraction, no helper, no utility module, no base class.
5. Only then: build the minimum that works. Smallest interface, deepest implementation. This is the Simplicity First principle: minimum code that solves the problem, nothing speculative. No features beyond what was asked. No abstractions for single-use code. No "flexibility" or "configurability" that wasn't requested.

Safety floor — never simplify away: trust-boundary validation, error handling, security, accessibility, data integrity. The ladder optimizes for less code, not less safety. If a rung would strip a safety guard, stop.

When a shortcut is deliberate, tag it so it's harvestable later:

# SHORTCUT: <what's skipped>. Upgrade: <what to do when this matters>.

Use grep -rnE '(#|//) ?SHORTCUT:' . across the repo to find accumulated shortcuts before they compound. See TDD for the shortcut-lifecycle workflow.

Ship and question: build the minimal version, then challenge the requirement. Does the caller actually need this shape? Could the seam move? If the requirement survives scrutiny, the minimal version was the right call. If it doesn't, you saved yourself the over-build. For adversarial requirement review, pair with Court Jester.

Glossary

Use these terms exactly — don't substitute "component," "service," "API," or "boundary." Consistent language is the whole point.

Module — anything with an interface and an implementation. Deliberately scale-agnostic: a function, class, package, or tier-spanning slice. Avoid: unit, component, service.

Interface — everything a caller must know to use the module correctly: the type signature, but also invariants, ordering constraints, error modes, required configuration, and performance characteristics. Avoid: API, signature (too narrow — they refer only to the type-level surface).

Implementation — what's inside a module, its body of code. Distinct from Adapter: a thing can be a small adapter with a large implementation (a Postgres repo) or a large adapter with a small implementation (an in-memory fake). Reach for "adapter" when the seam is the topic; "implementation" otherwise.

Internal seam — a seam private to a module's implementation, used by its own tests but never exposed through the external interface. A deep module can have internal seams (e.g., private helper classes, regional splits) without losing depth. Internal seams are invisible to callers; if they become visible, the module has lost its depth.

Depth — leverage at the interface: the amount of behaviour a caller (or test) can exercise per unit of interface they have to learn. A module is deep when a large amount of behaviour sits behind a small interface, shallow when the interface is nearly as complex as the implementation.

Seam (Michael Feathers) — a place where you can alter behaviour without editing in that place; the location at which a module's interface lives. Where to put the seam is its own design decision, distinct from what goes behind it. Avoid: boundary (overloaded with DDD's bounded context).

Adapter — a concrete thing that satisfies an interface at a seam. Describes role (what slot it fills), not substance (what's inside).

Leverage — what callers get from depth: more capability per unit of interface they learn. One implementation pays back across N call sites and M tests.

Locality — what maintainers get from depth: change, bugs, knowledge, and verification concentrate in one place rather than spreading across callers. Fix once, fixed everywhere.

Deep vs shallow

Deep module = small interface + lots of implementation:

┌─────────────────────┐
│   Small Interface   │  ← Few methods, simple params
├─────────────────────┤
│                     │
│  Deep Implementation│  ← Complex logic hidden
│                     │
└─────────────────────┘

Shallow module = large interface + little implementation (avoid):

┌─────────────────────────────────┐
│       Large Interface           │  ← Many methods, complex params
├─────────────────────────────────┤
│  Thin Implementation            │  ← Just passes through
└─────────────────────────────────┘

When designing an interface, ask:

• Can I reduce the number of methods?
• Can I simplify the parameters?
• Can I hide more complexity inside?

When shallow is correct. Depth is the default to reach for, not a law. A small, focused utility with little behind it is legitimately shallow — deepening it just to raise the depth ratio adds interface with nothing to hide. Deepen where complexity is escaping to callers; leave genuinely thin helpers thin.

Principles

• Depth is a property of the interface, not the implementation. A deep module can be internally composed of small, mockable, swappable parts — they just aren't part of the interface. A module can have internal seams (private to its implementation, used by its own tests) as well as the external seam at its interface.
• Deepening and internal splitting are orthogonal axes. Deepening shrinks the external interface (fewer modules, fewer imports). Internal splitting organizes the implementation (regions, private helpers, labeled sections). They serve different beneficiaries: deepening is a gift to the caller; internal splitting is a gift to the maintainer. They coexist when the interface stays small and stable. They conflict when internal structure leaks into the public API.
• The temporal sequence: gather first, then split. Consolidate fragmented logic into one module before decomposing internally. Splitting across files without consolidation is fragmentation wearing a disguise. Splitting within a consolidated domain is genuine decomposition.
• Internal splitting is legitimate when it serves human readability. It becomes an escape hatch when it's a cover for indecision ("we'll split internally for testability" while the class grows to 800 lines with no test at the interface level). The burden of proof: tests must exercise the public interface, not private methods.
• The 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.
• The interface is the test surface. Callers and tests cross the same seam. If you want to test past the interface, the module is probably the wrong shape.
• One adapter means a hypothetical seam. Two adapters means a real one. Don't introduce a seam unless something actually varies across it.

Designing for testability

Good interfaces make testing natural:

1. Accept dependencies, don't create them.

   # Testable
   def process_order(order, payment_gateway):
       ...
   
   # Hard to test
   def process_order(order):
       gateway = StripeGateway(api_key=os.environ["STRIPE_KEY"])
       ...
   

2. Return results, don't produce side effects.

   # Testable
   def calculate_discount(cart) -> Discount:
       ...
   
   # Hard to test
   def apply_discount(cart: Cart) -> None:
       cart.total -= discount
   

3. Small surface area. Fewer methods = fewer tests needed. Fewer params = simpler test setup.

Relationships

• A Module has exactly one Interface (the surface it presents to callers and tests).
• Depth is a property of a Module, measured against its Interface.
• A Seam is where a Module's Interface lives.
• An Adapter sits at a Seam and satisfies the Interface.
• Depth produces Leverage for callers and Locality for maintainers.

Rejected framings

• Depth as ratio of implementation-lines to interface-lines (Ousterhout): rewards padding the implementation. We use depth-as-leverage instead.
• "Interface" as the Python type hint or a class's public methods: too narrow — interface here includes every fact a caller must know.
• "Boundary": overloaded with DDD's bounded context. Say seam or interface.

Trigger Rules

Use these to decide when to redesign, not just how:

• Awkward change. When a feature feels awkward, one change spreads across files, or reviewers must reconstruct hidden dependencies, look for missing information hiding, shallow modules, temporal coupling, or complexity pushed to callers.
• New abstraction under scrutiny. When adding a module, layer, service, helper, wrapper, facade, option, callback, or argument, prove that it hides more complexity than it adds.
• API exposure check. When touching an API, check whether ordinary callers must know sequencing, representation, storage, transport, caching, protocol, file format, internal workflow, or too many setup steps.
• Special case pressure. When adding a special case, flag, exception path, conditional, or exposed container, first ask whether the owning module can eliminate the invalid state, isolate the unusual behavior, or provide a stronger operation.
• Split/extract audit. When splitting, extracting, or introducing variables, check whether the new boundary or name captures meaning or only adds jumps, pass-through state, and visible intermediate steps.
• Temporal coupling. When code is organized as prepare/process/finalize, staged objects, or other execution-order phases, verify that temporal structure is the real concept; otherwise reorganize around stable responsibilities.
• Naming quality. When naming is vague, mechanism-focused, inconsistent, or surprising, reconsider the abstraction boundary instead of accepting a near miss.
• Comment smell. When comments get long, duplicate code, justify a confusing interface, or explain usage by exposing internals, redesign the abstraction or move the missing contract to the interface.
• Performance pressure. When optimizing performance, measure first and hide the optimization; do not sacrifice module depth or information hiding without evidence that the tradeoff matters.

Comment Discipline

Comments reduce complexity when they document what the code cannot say about itself:

• Document: interface contracts, invariants, hidden design decisions, rationale, tricky implementation facts callers should not need to know.
• Do not narrate: code flow, obvious control flow, or restatements of the line above.
• Do not compensate for: bad names, poor decomposition, or confusing structure. Fix the code instead.
• When a comment explains why the interface is shaped this way (not what it does), keep it. That's design information the code can't carry.
• When a comment grows long to explain how to use an interface, the interface is too complex. Redesign it.
• Write the interface comment first. Sketch the public contract before implementing; if the comment is hard to write or keeps getting complicated, the abstraction is wrong — revise it before writing the code that would cement it.

Going deeper

• Deepening a cluster given its dependencies — see Deepening: dependency categories, seam discipline, and replace-don't-layer testing.
• Exploring alternative interfaces — see Design It Twice: spin up parallel sub-agents to design the interface several radically different ways, then compare on depth, locality, and seam placement.
• Platform-native reference — see Platform Native: what stdlib, frameworks, and browsers provide before you reach for a dependency.
• Complexity and comment discipline — see Complexity: cognitive load as primary metric, pull complexity downward, when to comment vs redesign.
• Dependency direction, module cycles, seam cost, and use-case separation — see Clean Architecture: inward dependencies, stability-directed module graphs, actor/change-reason separation, partial seams, and boundary testing.

Reference Index

See references/REFERENCE.md for the full routing table.