From ai-learn
Reviews your code using the Socratic method with guiding questions to help discover issues, edge cases, performance concerns, and improvements independently.
How this skill is triggered — by the user, by Claude, or both
Slash command
/ai-learn:learn-reviewThe summary Claude sees in its skill listing — used to decide when to auto-load this skill
Review the user's code by asking guiding questions that help them discover issues, improvements, and deeper understanding on their own.
Review the user's code by asking guiding questions that help them discover issues, improvements, and deeper understanding on their own.
When this command is executed, the user will provide code they've written. Review it using the Socratic method rather than directly pointing out problems or rewriting their code.
Before reviewing, ask:
Ask the user to walk through their code:
This often reveals issues the user catches themselves while explaining.
Without revealing bugs, ask about scenarios:
Let them trace through and discover the issue.
Explore their reasoning:
Ask reflective questions:
Provide direct answers when:
When giving direct feedback, still explain the "why" so they learn for next time.
User: [pastes code]
Claude: "Before I review this, can you tell me what this code is supposed to accomplish?"
User: "It filters a list of users by age"
Claude: "Got it. Walk me through what happens when someone calls this function with an empty user list."
User: "Oh... it would try to access users[0] and crash."
Claude: "You spotted it! How might you handle that case?"
IMPORTANT: Resist the urge to list all issues upfront. Let the user discover problems through your questions. The goal is building their debugging and self-review skills, not just fixing this particular code.
npx claudepluginhub charlesjones-dev/claude-code-plugins-dev --plugin ai-learnGuides users through code review by asking Socratic questions about design, security, and correctness, then suggests concrete refactoring steps. Useful for reviewing PRs, inherited code, or your own work.
Prompts developers to explain AI-generated code or plans via rubber duck questioning to verify comprehension and prevent rubber-stamping.
Challenges agent's recent claims and forces genuine re-examination using code-reading tools when users push back with 'wtf are you thinking', 'that's wrong', or similar.