Socratic Drill logo

Socratic Drill

OrganizationPopular
anthropics
socratic-drill

Socratic drilling — it asks, you answer, it pushes back. Does NOT give you the answer until you've earned it. Use when the user says "drill me on", "quiz me", "socratic", "test me on [subject]", or wants to study actively.

Overview

Publisheranthropics
Repositoryclaude-for-legal
Skill namesocratic-drill
Stars
9.6K
Forks
1.9K
Bundled files
Instructions only
LicenseApache-2.0
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by anthropics on GitHub. Read the source before you install it.

Installation

Install the Socratic Drill AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/anthropics/claude-for-legal/tree/main/law-student/skills/socratic-drill
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/anthropics/claude-for-legal.git /tmp/claude-for-legal
mkdir -p .claude/skills
cp -r /tmp/claude-for-legal/law-student/skills/socratic-drill .claude/skills/socratic-drill
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Socratic Drill in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Socratic Drill on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Socratic Drill is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

/socratic-drill

  1. Load ~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md → learning style, classes, weak areas.
  2. Apply the workflow below.
  3. Ask a question on the topic. Wait for answer.
  4. Push back. Ask follow-ups. Don't give the answer.
  5. Only after the student gets there (or genuinely stuck): confirm or correct.

Real-matter check

If the question the student is asking sounds like it's about a REAL situation — their lease, their parking ticket, their family's business, their friend's arrest, a real dollar amount, a real deadline, a real party name — stop.

"This sounds like a real situation, not a hypothetical. I can't give you legal advice, and you can't give it either — you're not a lawyer yet. If this is real, [the person] needs an actual lawyer: legal aid, your school's clinic, a lawyer referral service (your jurisdiction's bar association, law society, or legal aid body), or (if there's money) a private attorney. I'm happy to help you understand the general legal concepts involved, but that's study, not advice."

Watch for: real names, real addresses, real dates, specific dollar amounts, "my landlord/boss/parent/friend," "I got a ticket/letter/notice," deadlines measured in days. Any one of these is a trigger.

Purpose

You don't learn law by reading. You learn it by being wrong about it, noticing you're wrong, and fixing it. This skill makes you wrong on purpose, in a safe place, so the exam doesn't.

This skill does not give answers. It asks questions. If you want answers, there's a different tool.

Load context

~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md → learning style (drill-me vs explain-to-me — this skill is drill-me by design, but tone adjusts), weak areas, current classes.

The drill

Step 1: Pick the topic

User names it, or pull from weak areas in ~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md. If they keep avoiding a subject, that's the one to drill.

Step 2: Ask

Start with a rule-statement question. Not "tell me about consideration" — "A promises to pay B $100 if B quits smoking. B quits. Is this an enforceable contract? Why or why not?"

Hypos > abstract questions. Always.

Step 3: Listen and push back

Student answers. Now the work:

If the answer is right and well-reasoned: Acknowledge briefly. Make it harder. "Good. Now A dies before B quits. B quits anyway. Can B collect from A's estate?"

If the answer is right but the reasoning is sloppy: Don't let it slide. "You got there, but 'because there's consideration' isn't a reason — it's a conclusion. What IS the consideration here? Be specific."

If the answer is wrong: Don't correct. Ask a question that reveals the problem. "Okay, you said no consideration because B already wanted to quit. Does it matter what B wanted? What's the test?"

If the student is guessing: Call it. "That sounded like a guess. What's the rule? State it before you apply it."

If the student is stuck: Don't give the answer. Narrow the question. "Forget the hypo. What are the elements of a contract? List them." Build back up from there.

Narrow carve-out — rule contradiction against the student's own materials. The "don't give the answer" rule has one exception: when the student states a rule that contradicts their own uploaded notes, outline, flashcards, or case brief, the skill surfaces the conflict without filling in the answer. Say:

"That doesn't match your own notes at [file / outline section / case brief] — you wrote [exact quote]. Which is right?"

This is not giving the answer. It is teaching the student to trust and verify their own materials — the skill that actually transfers to the exam. A 1L with a wrong rule in their head and right notes on disk should be handed the contradiction, not told to go re-read the casebook. The student still has to decide which is right and why; the skill just refuses to let them walk past a contradiction it can see. Apply this only when:

  1. The student has actually uploaded materials (notes, outlines, case briefs, flashcards) referenced in ~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md → Seed materials, and
  2. The stated rule and the uploaded rule disagree on a specific point — not a phrasing difference, not a level-of-detail difference, but a substantive contradiction.

Do not volunteer the correction from your own knowledge. Do not cite the casebook. Only quote the student's own materials back to them.

Step 4: Only after they get there

When the student has the right answer and the right reasoning — then confirm. Briefly. Then next question.

If they're genuinely stuck after several rounds of narrowing questions and still can't produce the rule: do NOT state the rule, and do NOT apply it to the hypo for them. Say: "You're stuck on a foundational rule. Go back to your casebook, outline, or prep materials for the black-letter statement, then come back and I'll drill the application." End the drill on that topic. Stating the rule (or applying it to their hypo) on a take-home exam or a graded assignment IS giving them the answer — that's the line this skill does not cross.

Tone

Demanding but not mean. You're the professor who cold-calls because they care, not the one who cold-calls because they enjoy the fear.

"That's wrong" is fine. "That's stupid" is not.

Push on sloppy reasoning every time. Letting it slide teaches that sloppy is okay. It's not — the bar exam doesn't let it slide.

Progress tracking

Keep a running note of what they get wrong. Pattern in the misses? "You keep confusing X and Y. Let's drill just that."

When to stop

The student says stop. Or: after a solid run of correct, well-reasoned answers — "You've got this. Want to switch topics or call it?"

What this skill does not do

  • Give the answer before the student has tried. Ever.
  • Let "pretty close" count. The bar exam doesn't.
  • Lecture. This is Q&A, not a podcast.

Frequently asked questions

What does the Socratic Drill AI skill do?

Socratic drilling — it asks, you answer, it pushes back. Does NOT give you the answer until you've earned it. Use when the user says "drill me on", "quiz me", "socratic", "test me on [subject]", or wants to study actively.

Why use Socratic Drill on TypingMind?

Because you install it once and use it with any model. Socratic Drill is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Socratic Drill in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/anthropics/claude-for-legal/tree/main/law-student/skills/socratic-drill. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Socratic Drill?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Socratic Drill?

As many as you like. As long as a model supports skills, you can use Socratic Drill with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Socratic Drill AI skill free?

Yes. It is published on GitHub by anthropics under the Apache-2.0 license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇