Design Review

Design review of a codebase area, branch, or PR. Maps its bounded contexts, vertical slices, and modules, then judges where rules live and whether flexibility sits where change actually happens. Use when the user asks for a design or architecture review, or when a review should judge structure rather than bugs or style.

9 stars 2 forksGit & Code Review Plugin marketplace

Install

Claude Code plugin
/plugin marketplace add eralmansouri/skills

The repo is a Claude Code plugin marketplace. Run /plugin afterwards to pick the plugin that includes this skill.

skills CLI (Claude Code, Cursor, Codex and more)
npx skills add eralmansouri/skills --skill design-review
Manual install
git clone --depth 1 https://github.com/eralmansouri/skills.git /tmp/skills
mkdir -p ~/.claude/skills
cp -r /tmp/skills/skills/design-review ~/.claude/skills/design-review

Personal skills live in ~/.claude/skills. Use .claude/skills in a project folder to share a skill with your team.

From the SKILL.md

# Design Review

Review the **shape** of code: which context owns each rule, how use cases are cut into slices, which modules are deep or shallow, and whether the design bends where the domain actually changes. Bugs and style belong to a code review. This one judges design, and changes no code.

Two rules carry the review:

- **No map, no findings.** Map the design as it is, traced through real use cases, before judging any of it. A review without a map is a review of file names.
- **Evidence over taste.** Every finding cites code you opened, and every move it proposes passes the change test or the deletion test (see **Fit**). "This should use CQRS" is taste; "the refund window is checked in three places, with two different cut-offs" is evidence.

## Vocabulary

Domain nouns come from the project's `GLOSSARY.md` (one per context, or one at the root): "the Refund rules", not "RefundServiceImpl". Architecture nouns come from this table, one word per scale. Call the Skill tool with "codebase-design" for the rest of the module vocabulary (interface, depth, adapter, leverage, locality) and the deletion test.

| Say | Meaning | Instead of |
| --- | --- | --- |
| **Context** | A bounded context: a capability that owns its language, rules, and data | domain, service |
| **Slice** | One use case end to end, from entry point to persistence and side effects | endpoint, handler |
| **Module** | Anything with an interface and an implementation, at any scale | component, layer, service |
Read the full skill on GitHub

Details

Skill name
design-review
Path
skills/design-review/SKILL.md
Repo updated
15 October 2026

FAQ

What does the Design Review skill do?

Design review of a codebase area, branch, or PR. Maps its bounded contexts, vertical slices, and modules, then judges where rules live and whether flexibility sits where change actually happens. Use when the user asks for a design or architecture review, or when a review should judge structure rather than bugs or style.

How do I install the Design Review skill?

Claude Code plugin: /plugin marketplace add eralmansouri/skills. Skills go in ~/.claude/skills (personal) or .claude/skills (project), and Claude loads them automatically when a task matches the description.

Is Design Review popular?

It lives in eralmansouri/skills, which has 9 GitHub stars and 2 forks.

Related skills

Codebase Design

9

eralmansouri/skills

Use for code-level module design, deep modules, interfaces, seams, adapters, testability, refactoring shallow modules, and finding architecture improvements in an existing codebase.

Software Development

Domain Driven Design

9

eralmansouri/skills

Use for domain-driven design, domain modeling, ubiquitous language, bounded contexts, aggregates, invariants, entities, value objects, domain services, domain events, and deciding whether a rich domain model is warranted.

AI, LLMs & Agents

Event Driven Systems

9

eralmansouri/skills

Use for event-driven or message-driven architecture, commands and events, queues, pub/sub, Kafka, choreography, orchestration, sagas, outbox, idempotency, retries, ordering, backpressure, consumer lag, and messaging operations.

Software Development

Auto Experiment

5

zxl696/smallmodel-auto-experiment-skill

自动化超参数/架构搜索循环。给定模型代码、数据集和优化目标,自动生成 train.py,经双 agent 审计门禁通过后,循环调用 DeepSeek 提出改进建议、训练、用 git 追踪每轮变更,保留最优配置和 checkpoint。

Git & Code Review

Zephyr Explore

9

processmission/agent-skills

Use when Zephyr source, branch history, external precedent, or design constraints must be investigated before a decision can be made.

Git & Code Review

Cleanup Slops

9

comfuture/agent-skills

Find and simplify implicit state machines across a repository or a specified pull request without regressing behavior. Use for requests such as "find all of the implicit state machines in my app" or a cleanup pass that removes redundant state and consolidates scattered transitions for readability and maintenance.

Git & Code Review

Browse the most popular Claude skills, all git & code review skills, or write your own with the skills generator.