Skip to content

paskals/ai-assist

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

21 Commits
 
 
 
 
 
 
 
 

Repository files navigation

AI Coding Assistant Prompts

This repository is used to contain custom AI coding assistant prompts tailored for specific workflows and use cases.

Cline's Memory Bank

A modified version of Cline's memory bank instructions.

Custom Cline Memory Bank Instructions.

Claude Code Commands

Custom commands for Claude Code implementation workflows.

Planning

  • plan - Create comprehensive implementation specifications (specs) for requested changes.
  • review-spec - Review specification created by the plan command. Now includes optional independent verification via Codex reviewer for higher quality specs.

Implementation

  • implement - Implement features/changes based on existing spec files
  • review-implementation - Review the implementation of a spec. Critically evaluate if all required changes are implemented and if testing coverage is adequate. Now includes highly recommended independent verification via Codex reviewer to catch blind spots and security issues.

Claude Code Skills

Custom skills that extend Claude's capabilities.

Codex Independent Code Reviewer

A skill that provides independent code review by leveraging OpenAI's Codex CLI (GPT-5-Codex) as a second opinion alongside Claude's analysis.

Features:

  • Dual AI review (Claude + Codex) for higher confidence
  • Consensus highlighting when both find the same issues
  • Different perspectives revealing trade-offs and edge cases
  • Specialized workflows for security, performance, and architecture reviews

Installation:

cd claude-code/skills/codex-reviewer
chmod +x install.sh
./install.sh both  # Install for both Claude Desktop/Web and Code CLI

Quick Usage: Simply ask Claude to get a second opinion: "Review this code and get a second opinion from Codex"

Integration with Review Commands: The codex-reviewer skill is integrated into the review-spec and review-implementation commands to provide automatic dual-AI review:

  • After completing its analysis, Claude will invoke Codex for independent verification
  • Both perspectives are compared to identify consensus issues and unique findings
  • Review reports include an "Independent Review Findings" section
  • Significantly increases confidence in code quality and catches subtle bugs

See the full documentation for detailed usage, examples, and best practices.

About

Custom prompts and commands for AI assistants.

Resources

Stars

Watchers

Forks

Releases

No releases published

Contributors 2

  •  
  •