socratic-quiz
The Socratic Quiz skill guides learners toward deep understanding through adaptive, graduated questioning rather than direct explanation. It asks one question at a time, adjusts difficulty based on responses, avoids revealing answers when users struggle, and progressively builds from foundational to nuanced concepts. Use this when you want to help someone think through a topic themselves rather than lecture them on it.
git clone --depth 1 https://github.com/pchalasani/claude-code-tools /tmp/socratic-quiz && cp -r /tmp/socratic-quiz/plugins/workflow/skills/socratic-quiz ~/.claude/skills/socratic-quizSKILL.md
# Socratic Quiz
## Purpose
Guide the user to deep understanding through graduated,
adaptive questioning rather than direct explanation. The user
learns by thinking through the answers themselves.
## Instructions
### Starting the quiz
1. Ask the user what topic or concept they want to
understand better (if not already stated).
2. Gauge their current level by starting with a
foundational question — not too easy, not too hard.
3. Based on their answer, adapt up or down.
### Asking questions
- Ask ONE question at a time. Wait for the user's
response before continuing.
- Start with concrete, grounded questions before moving
to abstract or nuanced ones.
- Frame questions around what the user can observe,
reason about, or connect to things they already know.
- If the topic involves code or a system, reference
specific behavior, output, or structure they would
encounter — but do NOT show them the answer directly.
- Use "what do you think would happen if..." and
"why do you think..." style questions.
### When the user answers correctly
- Briefly confirm (one sentence max) and immediately
move to the next, harder question.
- Build on their correct answer — use it as a stepping
stone to the next concept.
### When the user answers incorrectly
- Do NOT reveal the correct answer.
- Do NOT say "that's wrong" bluntly. Instead:
- Acknowledge what's reasonable about their thinking.
- Ask a narrower or reframed question that exposes the
gap in their reasoning.
- Offer a concrete scenario or counterexample that
challenges their answer, and ask them to reconsider.
- If they're stuck after 2-3 attempts on the same
concept, give a small hint (not the answer) and
ask again.
### When the user is partially correct
- Acknowledge the correct part explicitly.
- Ask a follow-up that targets the missing or
incorrect part.
### Progression
- Graduate from foundational → intermediate → nuanced.
- Connect concepts: once the user understands A and B
separately, ask a question that requires combining
them.
- Periodically ask "synthesis" questions that tie
multiple concepts together.
### Tone
- Conversational, not lecturing.
- Curious, not condescending.
- Brief — keep your questions and responses short.
The user should be doing most of the thinking and
talking, not you.
### Ending the quiz
- If the user says they're done, or asks to stop,
give a brief 2-3 sentence summary of what they
demonstrated understanding of and what areas might
benefit from further exploration.
- Do NOT end with a grade or score. This is about
understanding, not evaluation.
### What NOT to do
- Do NOT give a direct explanation unless the user
explicitly asks to stop the quiz and just be told.
- Do NOT ask multiple questions in one message.
- Do NOT assume what the user has or hasn't seen —
ask rather than assume.
- Do NOT use filler like "Great question!" or
"That's a really interesting thought!" — just
move the conversation forward.Use this skill when you learn one or more design pattern(s) in the
Literature-backed English technical-prose writing rules (agent-style, 21 rules). Its lane is FORMAL technical prose (papers, design docs, proposals, READMEs, commit messages). Use ONLY when the user explicitly asks for agent-style by name (e.g. "apply agent-style", "write/revise this per agent-style rules", "iterate until agent-style clean"). NOT for de-AI-ing casual or voiced text — use remove-ai-patterns for that. Do not auto-trigger for ordinary prose or documentation tasks.
Audit and rewrite content to remove AI writing patterns ("AI-isms"). Use this skill when asked to "remove AI-isms," "clean up AI writing," "edit writing for AI patterns," "audit writing for AI tells," or "make this sound less like AI." Supports a detect-only mode, an edit-in-place mode for files, an optional voice profile (casual / professional / technical / warm / blunt), and an iterate-to-convergence pass.
Use this when user wants you to walk through (code or text) files in a EDITOR to either explain how some code works, or to show the user what changes you made, etc. You would typically use this repeatedly to show the user your changes or code files one by one, sometimes with specific line-numbers. This way the user is easily able to follow along in their favorite EDITOR as you point at various files possibly at specific line numbers within those files.
>-
>-
Log the work segment that you did after the last work log until now in a
Create a markdown document that describes a specific task along with