OpenSkillIndex← back to search
claude-code

Awesome-omni-skill code-quality

A protocol for performing a deep code quality analysis using static analysis tools and formatting. It covers C++ (cppcheck) and React/JavaScript (ESLint).

install
source · Clone the upstream repo
git clone https://github.com/diegosouzapw/awesome-omni-skill
Claude Code · Install into ~/.claude/skills/
T=$(mktemp -d) && git clone --depth=1 https://github.com/diegosouzapw/awesome-omni-skill "$T" && mkdir -p ~/.claude/skills && cp -r "$T/skills/development/code-quality" ~/.claude/skills/diegosouzapw-awesome-omni-skill-code-quality-af8435 && rm -rf "$T"
manifest: skills/development/code-quality/SKILL.md
source content

Agentic Command: /check

When the user writes 

/check
in chat, immediately trigger the full code-quality protocol below, without further confirmation. This command acts as a shortcut for a comprehensive code review and static analysis. The agent should proceed automatically, applying all steps unless the user specifies otherwise.

Agentic Code Analysis and Quality Protocol

As an AI assistant, your goal is to perform a comprehensive code quality check using industry-standard tools. When asked to "review the code," "check for bugs," or "improve code quality," follow this protocol.

Step 1: Identify File Type and Scope

  1. Determine the language of the file(s) to be reviewed:
    • C++ (
      .cpp
      , 
      .h
      )
    • React/JavaScript/CSS (
      .jsx
      , 
      .js
      , 
      .scss
      )
  2. Ask the user if you should apply formatting changes directly or just provide suggestions.

Step 2: Run Static Analysis

This is the core of the code review. Run the appropriate tool to find bugs, vulnerabilities, and code smells.

For C++ Firmware (
src/
directory):

  1. Run Static Analysis: Execute the PlatformIO 
    check
    command. 
    platformio check
  2. Analyze the output for any reported defects, vulnerabilities, or performance issues.

For React/JavaScript Frontend (
web/
directory):

  1. Check for ESLint setup:
    • Look for an 
      .eslintrc.*
      file in the root or 
      web/
      directory.
    • Check 
      web/package.json
      for a 
      lint
      script.
  2. If ESLint is not configured:
    • Inform the user that ESLint is not set up.
    • Ask for permission to set it up: "I've noticed ESLint isn't configured for the frontend code. It's a powerful tool for finding bugs. Would you like me to set it up for you?"
    • If yes, proceed to install 
      eslint
      and a standard configuration (e.g., 
      eslint-plugin-react
      ). Create a basic 
      .eslintrc.js
      and add a 
      lint
      script to 
      package.json
      .
  3. If ESLint is configured:
    • Execute the linting command. 
      npm run lint --prefix web
    • Analyze the output for any errors or warnings.

Step 3: Apply Automatic Formatting

After the deep analysis, apply standard code formatting for consistency.

For C++ Firmware:

  1. Formatting: If the project provides a formatting script (e.g., 
    scripts/format.sh
    ) or a documented process, use that. Otherwise, use 
    clang-format
    directly: 
    clang-format -i src/**/*.cpp src/**/*.h
    (Adjust the file pattern as needed for your project structure.)
  2. If no formatting tool is available, inform the user and offer to help set up 
    clang-format
    with a standard configuration file (e.g., 
    .clang-format
    ).

For React/JavaScript Frontend:

npm run format --prefix web

(Note: If the 

format
script doesn't exist, offer to set it up with Prettier.)

Step 4: Report Findings and Suggestions

  1. Summarize Static Analysis Results: Report the critical findings from 
    cppcheck
    or 
    ESLint
    . For each issue, provide the file, line number, and a description of the problem.
  2. Summarize Formatting: Mention that the code has been automatically formatted for consistency.
  3. Offer to Fix: Ask the user if they would like you to attempt to fix the issues found by the static analysis tools.