How does this differ from a standard linter?

While linters focus on micro-style and syntax errors, this skill focuses on macro-architectural patterns such as module responsibility, error propagation, and overall structural reliability.

Can it help improve TypeScript safety?

Yes, it specifically audits for unsafe type assertions, optional chaining overuse that hides null sources, and unchecked regex matches, suggesting type-safe alternatives.

Does it provide specific refactoring steps?

Yes, it provides concrete recommendations including proposed file names, line numbers, and specific structural changes rather than just general observations.

What does it mean for a codebase to be 'LLM-friendly'?

It refers to how easily an AI assistant can interpret the code, evaluated through documentation coverage, descriptive naming, and clear error messages that explain intent.

Architecture Reviewer

Name: Architecture Reviewer
Author: getsentry

bygetsentry

•

134

•

Security & Testing

Performs staff-level architectural audits to identify structural technical debt, reliability gaps, and maintenance bottlenecks.

This skill transforms Claude into a staff engineer capable of performing deep, macro-level architectural audits on any codebase. Instead of focusing on minor style preferences, it identifies systemic issues that compound over time, such as monolithic modules, silent error handling patterns, and type safety gaps. It evaluates the codebase across five critical dimensions—complexity, failure patterns, type safety, test coverage, and AI-readiness—providing prioritized recommendations to improve long-term maintainability and system reliability.

Key Features

01Detects silent failure patterns like swallowed errors and empty return statements.

02Audits TypeScript safety gaps including unsafe 'as' casts and unchecked regex matches.

03134 GitHub stars

04Identifies monolithic files over 500 lines and proposes logical module splits.

05Maps test coverage against critical business logic to identify high-risk gaps.

06Assesses LLM-friendliness by evaluating JSDoc quality and naming clarity.

Use Cases

01Auditing a new repository to identify hidden technical debt and structural risks.

02Ensuring critical error paths are properly surfaced and tested in high-reliability systems.

03Performing a comprehensive health check on a legacy codebase before a major refactor.

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add getsentry/warden architecture-review

For use in Claude.ai and ChatGPT

Download Skill