Skip to content
Back to skills

Codebase Design

ASecurity

Use for substantive module-interface, encapsulation, or testability decisions. Skip ordinary code explanations and cosmetic refactors.

  • 2 stars
  • 0 votes
  • 0 copies
  • 3 views
  • Added October 15, 2026
designtestinggitapiperformance

Works with

  • api

Security analysis

A100/100

Pro scans all 4 files and shows the line behind each finding

Scanned October 15, 2026

npx -y skills add MVPavan/via --skill codebase-design --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Codebase Design?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Codebase Design
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/mvpavan-codebase-design/badge)](https://www.skillsdirectory.com/skills/mvpavan-codebase-design)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
name: codebase-design
description: Use for substantive module-interface, encapsulation, or testability decisions. Skip ordinary code explanations and cosmetic refactors.
---

# Codebase Design

Design interfaces that hide useful complexity and concentrate change. Apply this
lens to real interface or module decisions, not every explanation of code.

## Vocabulary

- **Module:** an interface and its implementation, at a useful scale.
- **Interface:** what callers must know, including invariants, ordering, errors,
  configuration and performance constraints, not just type signatures.
- **Depth:** useful behavior hidden behind a manageable interface.
- **Seam:** a place where behavior can vary without changing its callers.
- **Adapter:** a concrete implementation filling a role at that seam.
- **Locality:** related knowledge, changes and checks stay together.

Use established project vocabulary, including API, service or boundary where
precise. These definitions aid reasoning rather than replacing the domain glossary.

## Design checks

Ask what callers gain, what complexity disappears from them, and what knowledge
still leaks through. If deleting a layer merely removes forwarding, question its
value; if it spreads complexity across callers, the layer may be earning its keep.

An interface can be justified by encapsulation, ownership or a present testing
need even with one implementation. Additional adapters are evidence of variation,
not a universal prerequisite. Internal seams and focused internal tests are valid;
do not expose internals merely to satisfy a testing rule.

Prefer explicit dependencies and pure calculations where they clarify behavior;
side effects are legitimate responsibilities when isolated and well specified.
Preserve compatibility and operational constraints when deepening a module.

For dependency-oriented deepening, use `DEEPENING.md`. For a consequential interface
choice with multiple credible designs, optionally use `DESIGN-IT-TWICE.md`;
ordinary design decisions do not require multiple agents or variants.

Files in this skill

  • DEEPENING.md1.1 KB
  • DESIGN-IT-TWICE.md720 B
  • SKILL.md2 KB
  • agents/openai.yaml42 B

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.