Can it detect logic errors that linters miss?

Yes, it specifically looks for conceptual issues like incorrect boolean logic, state management bugs, and missed edge cases that standard static analysis tools cannot catch.

When should I use this instead of a specialized review skill?

Use this as your default review type. It is ideal for general quality, logic, and readability. For specific audits like security or performance profiling, use more specialized plugins.

How does it handle project-specific coding styles?

The skill begins with context research, scanning for files like CLAUDE.md, .editorconfig, and linter settings to ensure its feedback aligns with your project's unique rules.

What does the standard-review Claude Code skill do?

It performs a comprehensive 'senior developer' pass on your code, checking for logic bugs, readability issues, and consistency with your project's specific standards.

Standard Code Review

Name: Standard Code Review
Author: liminal-ai

byliminal-ai

0•

Collaboration Tools

Conducts comprehensive, senior-level code reviews focusing on logic correctness, readability, and adherence to project conventions.

The standard-review skill transforms Claude into an experienced senior developer, providing a thorough 'second pair of eyes' for your codebase. It follows a rigorous three-phase methodology—Context Research, Comparative Analysis, and Issue Assessment—to ensure that feedback is not just generic advice, but deeply relevant to your project's existing patterns and configuration files. Whether you are preparing a Pull Request or looking for a quick sanity check on a new implementation, this skill identifies subtle logic bugs, naming inconsistencies, and architectural deviations that automated linters often miss.

Key Features

01Function-level assessment of single responsibility and code duplication

02Deep logic bug detection for edge cases, off-by-one errors, and state management

03Strict adherence checks for project conventions like CLAUDE.md and editor configurations

040 GitHub stars

05Multi-phase review process including context research and pattern identification

06Readability and naming clarity analysis to reduce future maintenance debt

Use Cases

01Performing a final senior-level check on Pull Requests before merging

02Identifying 'silent' logic errors and unhandled edge cases in complex functions

03Auditing new implementations for consistency with existing project patterns

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add liminal-ai/code-steward standard-review

For use in Claude.ai and ChatGPT

Key Features

01Function-level assessment of single responsibility and code duplication

02Deep logic bug detection for edge cases, off-by-one errors, and state management

03Strict adherence checks for project conventions like CLAUDE.md and editor configurations

040 GitHub stars

05Multi-phase review process including context research and pattern identification

06Readability and naming clarity analysis to reduce future maintenance debt

Use Cases

01Performing a final senior-level check on Pull Requests before merging

02Identifying 'silent' logic errors and unhandled edge cases in complex functions

03Auditing new implementations for consistency with existing project patterns