husamql3/arch-review-skill/arch-review

Claude Code and agent skill by husamql3/arch-review-skill

Trust tier 2
  • community
LLM review pending
Source check pending

Checking that the SKILL.md file is still in the repository.

Creates a visual architecture review as a before/after HTML walkthrough of code changes, including modules, dependencies, seams, function signatures, and rationale. Use when the user asks to walk through architecture changes, explain before/after design, review a refactor, or visualize module dependencies with Mermaid.

Install

npx vskill@latest install husamql3/arch-review-skill/arch-review

SKILL.md

Arch Review

Create a standalone HTML file that explains architectural changes from before to after. The artifact should help a reviewer understand what moved, why it moved, and where the new seams are.

Workflow

  1. Establish the comparison base.
  • If the user names a branch, commit, PR, diff, or file set, use that.
  • Otherwise inspect git state and choose the safest obvious base: merge-base with the main branch, the previous commit, or unstaged changes.
  • Completion criterion: every changed file in scope has a clear before/after source.
  1. Inventory the architecture.
  • Identify modules, packages, entry points, exported types, public functions, adapters, boundaries, and cross-module dependencies.
  • Prefer static analysis tools already present in the repo; otherwise use rg, language-aware parsers, import graphs, and targeted file reads.
  • Completion criterion: every changed module and every direct dependency crossing its boundary is accounted for.
  1. Find the seams.
  • Mark seams where policy meets IO, domain meets framework, sync meets async, external service adapters begin, or ownership boundaries change.
  • For each seam, record the old shape, new shape, and who calls whom.
  • Completion criterion: every new or removed boundary has an explanation tied to concrete files and symbols.
  1. Capture signatures.
  • List changed public function, class, method, component, hook, API route, command, event, or schema signatures.
  • Include before and after forms when they differ.
  • Completion criterion: every caller-visible contract change is represented.
  1. Produce the HTML walkthrough.
  • Save a single self-contained .html file unless the user asks otherwise.
  • Include sections for summary, before architecture, after architecture, dependency changes, seams, signature changes, migration impact, risks, and open questions.
  • Use Mermaid diagrams for module/dependency flows and sequence diagrams when call flow changed.
  • Completion criterion: the file opens locally and the diagrams render from the included Mermaid CDN script.
  1. Verify the artifact.
  • Open or render the HTML when browser tooling is available.
  • Check that diagrams are non-empty, labels fit, and changed files/symbols are traceable.
  • Completion criterion: report the saved file path and any verification gaps.

HTML Requirements

  • Use a readable technical-report layout, not a landing page.
  • Include a table of contents.
  • Use file paths and symbol names as anchors where useful.
  • Prefer Mermaid for dependency graphs:
  • flowchart LR for module dependencies.
  • sequenceDiagram for runtime flows.
  • classDiagram only when types/classes are central to the change.
  • Keep Mermaid labels short; put details in nearby tables.
  • Include a small legend for edge meanings such as calls, imports, emits, persists, or adapts.
  • Do not claim intent that cannot be inferred from code or user context. Mark it as an inference or open question.

Question Discipline

Ask the user one question at a time only when code inspection cannot answer it. Provide a recommended answer with each question. Do not implement additional architecture changes unless the user explicitly asks.

Output Contract

The final response must include:

  • The path to the generated HTML file.
  • The comparison base used.
  • A brief note on verification performed.
  • Any open architectural questions left unresolved.

Security

  • Security scanPassed, 100 of 100Pattern checks on the skill's files.
  • Model reviewNot runA model review is scheduled for the next rescan.

Quality evaluation

  • Task evaluationNot runNo evaluation yet. Run one in Skill Studio.

Source

Popularity

7-day trend score
0

Not refreshed yet.

Works with

Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.

  • Claude Code
  • Cursor
  • GitHub Copilot
  • Windsurf
  • Codex
  • Gemini CLI
  • Cline
  • Amp
  • Roo Code
  • Goose
  • Continue
  • Kilo Code
  • Junie
  • Trae
  • OpenHands
  • Qwen Code
  • AI Agent
  • Tabnine Chat
  • Sourcegraph Cody
  • Amazon Q
  • Firebase Genkit
  • Sweep
  • Devin
  • Phind
  • Mentat
  • Claude Desktop
  • Aider
  • GPT Pilot
  • Plandex
  • bolt.new
  • v0
  • Replit Agent
  • Anse
  • TypingMind
  • BetterChatGPT
  • LibreChat
  • Warp AI
  • Pieces
  • Fig AI
Embed badge
vskill VERIFIED badge
[![vskill VERIFIED](https://verified-skill.com/api/v1/skills/husamql3/arch-review-skill/arch-review/badge)](https://verified-skill.com/skills/husamql3/arch-review-skill/arch-review)

Something wrong with this skill or its page? Tell us.