Aluisio Amorim

You can critically review implementation plans before execution, get senior staff engineer level code reviews covering bugs, security, performance, and architecture, and author, edit, or verify new skills before deployment. Each skill spawns a Claude Code subprocess to perform the analysis.

These skills target software engineers and applied AI practitioners who want pre-execution plan validation, rigorous code review from a senior engineering perspective, and a structured process for building and verifying their own skills before shipping them.

Both skills spawn a Claude Code subprocess when invoked. review-plan analyzes an implementation plan for inconsistencies, missing edge cases, and architectural concerns, while review-code analyzes code for bugs, security issues, performance problems, and architectural concerns.

You need a Claude Code environment capable of spawning subprocesses, since each skill delegates its analysis to a Claude Code subprocess. No other dependencies, services, or external integrations are specified in the skill manifest.

Use writing-skills whenever you are creating new skills, editing existing skills, or verifying that skills work correctly before deployment. It governs the skill authoring lifecycle rather than plan or code analysis.