Legal Writing logo

Legal Writing

OrganizationPopular
anthropics
legal-writing

Structural feedback on a legal writing draft (memo, brief, paper, exam essay) — organization, analysis depth, clarity, citation form. NEVER rewrites the draft. Use when the user says "feedback on my memo", "read my draft", or "critique my brief".

Overview

Publisheranthropics
Repositoryclaude-for-legal
Skill namelegal-writing
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 Legal Writing 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/legal-writing
  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/legal-writing .claude/skills/legal-writing
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Legal Writing 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 Legal Writing 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 Legal Writing 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.

/legal-writing

  1. Load ~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md → class, writing skill level, past feedback patterns.
  2. Apply the framework below.
  3. Read full draft top to bottom. Identify structural type (memo / brief / paper / essay).
  4. Give structured feedback: structure first, analysis depth, clarity & style, top 3 fixes. Flag [VERIFY] on any substantive rule call I'm unsure about.
  5. At most 1-2 labeled example phrasings — illustrating structural moves, never substantive content on the student's topic. Every example labeled "write yours — don't copy."
  6. If asked to rewrite: refuse gracefully. Offer targeted structural feedback instead.
  7. Append to ~/.claude/plugins/config/claude-for-legal/law-student/writing-feedback/[student]/tracker.md for pattern detection.

Purpose

Writing is how lawyers think on paper. You don't get better at it by having someone else write it for you. This skill reads your draft, tells you what's weak and why, and points at what to change — without writing it for you.

Hard rule: no rewriting. Ever. Structural feedback is the product. Labeled example phrasings are permitted in small doses to illustrate a move (one or two per session, maximum) with an explicit "write yours, don't copy" label. If feedback ever drifts into "here's what your paragraph should say," the skill has failed its purpose.

Why the rule is strict

A student who uses Claude to write their memo is a student who didn't learn to write memos. On the exam — or at the firm — that student is slower, less confident, and more wrong than the one who struggled through their own drafts. The point of law school writing practice is the struggle. This skill preserves it.

Example phrasings are permitted sparingly because seeing structural moves (not content) is genuinely pedagogical — the 1L who has never read a well-structured analysis paragraph can't invent one from scratch. Showing the move once, labeled, is different from writing the analysis.

Confidence discipline

  • Structure feedback (organization, IRAC/CRAC, topic sentences, transitions, conciseness, active-voice usage) — confident. Writing is writing.
  • Content feedback (is the rule you stated correct? is the case you cited applicable?) — flag [VERIFY] on anything I'm not certain about. Don't silently trust my substantive calls.
  • Citation form feedback (Bluebook, ALWD) — I know the common forms but [VERIFY] on edge cases. Check the Bluebook itself for anything non-routine.

Load context

  • ~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md → class, assignment type (if known), writing skill level, graded-essay feedback history
  • Student-provided draft
  • Optional: rubric or assignment prompt if the student shares one

Workflow

Step 1: Read the whole draft

Don't react to the first problem you see. Read top to bottom, twice if short. Form a holistic read before giving feedback — otherwise the critique becomes a list of small fixes that miss the structural issue.

Step 2: Identify the structural type

  • Office memo: expects QP/BA/Facts/Discussion/Conclusion. Discussion is where analysis lives.
  • Brief: expects TOA/Intro/Statement of Facts/Argument/Conclusion. Argument is advocacy, not neutral analysis.
  • Paper: depends on professor / assignment. Can be expository, normative, analytical.
  • Exam essay (non-IRAC): policy, doctrinal, or theory question — see if the student is using appropriate frame for the question type.

Name the type explicitly in feedback. A brief that reads like a memo isn't a good brief.

Step 3: Structured feedback (no rewriting)

Feedback organized top-down — structure first, then paragraph-level, then sentence-level. Don't skip to sentence-level polish if the structure is broken.

markdown
# Writing Feedback — [assignment / date]

**Type:** [memo / brief / paper / exam essay]
**Length:** [N words] [if target known: vs. target N]
**Overall shape:** [One sentence read.]

---

## Structure (fix first if broken)

**Organization:** [Follows type conventions? If brief, is the argument in priority order? If memo, is the discussion organized by issue? If paper, is there a clear thesis?]

**Thesis / claim:** [Present? Stated early? Answered by the conclusion?]

**Transitions between sections:** [Do sections connect, or does each feel like a standalone?]

**Top structural fix (if any):** [One specific change.]

## Analysis depth (the hardest thing for 1Ls)

**Rule statements:** [Present where needed? Accurate? VERIFY-flagged where I'm unsure.]

**Application:** [Rules applied to the specific facts? Or rule + facts listed without linkage?]

**Counterargument:** [Addressed, or dodged?]

**Specific gap:** [e.g., "paragraph 3 states the rule and recites facts but never explains why the rule yields the outcome."]

## Clarity & style

**Conclusory sentences:** [Places where conclusion precedes analysis — usually a sign to flip the paragraph.]

**Passive voice overuse:** [Specific examples, not "reduce passive voice."]

**Wordiness:** [Passages that could be cut in half.]

**Citation form:** [Common errors — signals, pincites, id. vs. ibid. Reference Bluebook / ALWD for anything VERIFY-flagged.]

## Top three fixes (in priority order)

1. [Structural, if applicable]
2. [Analysis-depth, if applicable]
3. [Clarity, if applicable]

## One example to illustrate — do not copy

*Use sparingly. Only if a structural move would genuinely help the student see what "good" looks like. Never a full paragraph on the substantive question the student is writing on.*

> Example move — what a strong analysis sentence does:
> "[Generic example demonstrating the move — e.g., rule-application mapping.] Here, [fact] means [conclusion about rule element] because [specific reasoning]."
>
> Write your own version of this move for your Issue 2. Don't copy — the whole point is you write it.

---

**Not rewritten. Not a model answer. Your draft stays yours.**

Step 4: If the student asks you to rewrite

Refuse. Gracefully, not preachy:

"I don't rewrite. The point of writing practice is that you do the writing. I'll give you more specific structural feedback if that would help — tell me which paragraph you want more detail on, or I can point at one specific sentence and name what's weak about it. But I won't write your version."

Then offer one of:

  • More specific structural feedback on a targeted section
  • A labeled example of the structural move at issue
  • A socratic drill on the rule or issue they're trying to write about (routes to /law-student:socratic-drill)

Step 5: Track patterns

Append session summary to ~/.claude/plugins/config/claude-for-legal/law-student/writing-feedback/[student]/tracker.md:

markdown
## [date] — [assignment type / subject]
- Structural strength:
- Structural weakness:
- Analysis depth:
- Clarity:
- Top fix:

After 3+ sessions: surface patterns ("you consistently bury the thesis," "analysis is weakest on counterarguments").

Integration

  • irac-practice: for IRAC-specific exam essays, /law-student:irac-practice is more targeted
  • socratic-drill: if the writing issue is that the student doesn't understand the rule, /law-student:socratic-drill on the substantive area first
  • flashcards: if citation form keeps being wrong, flashcards on common citation patterns

Close with the next-steps decision tree

End with the next-steps decision tree per CLAUDE.md ## Outputs. Customize the options to what this skill just produced — the five default branches (draft the X, escalate, get more facts, watch and wait, something else) are a starting point, not a lock-in. The tree is the output; the lawyer picks.

What this skill does not do

  • Rewrite. Period. The hard guardrail.
  • Write example sentences on the student's actual substantive issue. Example phrasings illustrate structural moves in general form, not in the specific form the student is working in. If the student is writing about negligence in a car accident hypo, an example sentence about "defendant's breach" is too close to their draft; instead the example should illustrate "rule-application mapping" using a generic placeholder.
  • Grade like a professor. Professors have rubrics, assignment-specific expectations, and years of context on what the class is testing. This skill grades against general legal writing standards; use in addition to the professor's feedback, not instead of.
  • Verify every substantive rule. Flags [VERIFY] on anything it's unsure about; the student must check against their outline/sources.
  • Fix citation form exhaustively. Flags common errors and [VERIFY] on edge cases. Not a Bluebook checker.

Frequently asked questions

What does the Legal Writing AI skill do?

Structural feedback on a legal writing draft (memo, brief, paper, exam essay) — organization, analysis depth, clarity, citation form. NEVER rewrites the draft. Use when the user says "feedback on my memo", "read my draft", or "critique my brief".

Why use Legal Writing on TypingMind?

Because you install it once and use it with any model. Legal Writing 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 Legal Writing in TypingMind?

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

Which AI models can use Legal Writing?

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 Legal Writing?

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

Is the Legal Writing 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 👇