Live survey304 models · 14 providers · 22,792 auditsAudited grill-with-docs · SAFE →What changedHow we verifyllms.txt

Agent skill · Audited from source

cold-call-prep

skill · anthropics/claude-for-legalApache-2.0 licenceAudited 31 Oct 2026 · source cb488a888d20

★ 9,858pushed 2 days ago
✓SAFEGrade B · 89/100 · 31 Oct 2026 · cb488a8
9,858repo stars · whole anthropics/claude-for-legal

Is it safe? Yes — SAFE, grade B (89/100). We read this skill's source at commit cb488a888d20 and found no filesystem, network or shell access and no hard-coded secrets. What it does: Prep for a cold-call — predict the professor's likely questions and drill them Socratically, flagging where you're shaky so you know what to re-read before class. Use when the user says "prep for class tomorrow", "cold…

Ask an AI about this pageGChatGPTCClaudePPerplexityGeGeminiXGrok
More
Cite

Cite this page

ModelsAtlas. “cold-call-prep (anthropics/claude-for-legal)”. Retrieved 31 Oct 2026. https://modelsatlas.com/skills/anthropics/cold-call-prep

Trust audit

Audit 0.4.1 · source read 31 Oct 2026

Trust score89/100✓SAFE B
0 critical0 high0 medium0 low0 info

SAFE means nothing in the source contradicts what the skill says it does. Grade B is shared by 16,667 of 17,602 audited skills.

What the source does

  • FilesystemNone observedNo reads or writes in the source
  • NetworkNone observedNo outbound calls in the source
  • ShellNone observedNo commands run in the source
  • DependenciesPinnedNothing floats
  • SecretsNone foundNo keys in source

Five audit layers

LayerWhat we checkResultFindings
L0 ProvenanceRepo owner, licence, archive status, activityPASS0
L1 Static sourceSecrets, path traversal, supply chain, hidden codeNA0
L2 InstructionsPrompt injection, authority framing, hidden directivesPASS0
L3 InventoryThe files and declarations match the sourcePASS0
L4 SandboxRun in isolation and watch behaviourSKIPPED why0

Gates applied: no behavioural pass.

Findings (0)

No findings at any severity. We re-audit when the source changes.

Audit 0.4.1 · source sha cb488a888d20 · 31 Oct 2026Report an issue or request a re-scan

Provenance

Who published it, and what the verdict is pinned to

  • Publisheranthropics GitHub organizationOwner of the repository this skill is read from
  • Source folderanthropics/claude-for-legal/law-student/skills/cold-call-prepThe exact folder we audited
  • LicenceApache-2.0As the repository states it
  • Commit pinned for this verdictcb488a888d20A new audit runs when the source changes
  • Repository statusNot archivedLast push 2 days ago
  • Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

What it tells the agent

Full source on GitHub ↗

Its description: Prep for a cold-call — predict the professor's likely questions and drill them Socratically, flagging where you're shaky so you know what to re-read before class. Use when the user says "prep for class tomorrow", "cold…

---
name: cold-call-prep
description: >
  Prep for a cold-call — predict the professor's likely questions and drill
  them Socratically, flagging where you're shaky so you know what to re-read
  before class. Use when the user says "prep for class tomorrow", "cold call
  [case]", "what might [professor] ask on", or points at assigned reading.
argument-hint: "[case name, or paste case text, or path to reading]"
---

# /cold-call-prep

1. Load `~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md` → class list, professors, learning style.
2. Apply the workflow below.
3. Identify reading (case name + citation, professor, class, syllabus context).
4. Predict 6-10 likely questions across categories (Facts / Holding / Reasoning / Application / Policy), weighted to professor's known tendencies.
5. Drill using socratic pattern — ask, wait, push back, narrow when stuck. Don't give answers.
6. Post-drill summary: strong/shaky/missed; what to re-check before class.

---

## 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

Cold-calling lives or dies on preparation. The professor has read the case dozens of times and knows the questions; the student has read it once. This skill narrows the gap — predicts the likely question patterns for the case, drills the student on them, and surfaces what they haven't locked in.

Not a replacement for reading the case. A test that you actually did.

## Confidence discipline

- When the student provides case text or casebook excerpts: I predict questions based on the actual text. Confident.
- When the student provides only a case name: I predict based on what I know about the case. Flag `[UNCERTAIN]` on any question that depends on case details I'm not sure of. Strongly recommend the student pastes the case or casebook treatment first.
- If I don't know the case well: say so. "I don't have a reliable read on this case — paste the text or casebook treatment and I can work from that. Otherwise my questions are educated guesses."

## Load context

- `~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md` → current classes, professors, learning style
- User-provided: case name / case text / casebook pages / reading list

## Workflow

### Step 1: Identify the reading + professor

- Case name and citation
- Professor (from ~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md class list — tone and focus vary by professor)
- Class / subject area
- Where this case falls in the syllabus (for context — is this the first case on the topic, a narrowing case, a counterexample?)

### Step 2: Predict the questions

Professors cold-call in recurring patterns. Predict across these categories:

**Facts-level (warm-up):**
- Who are the parties? What happened? Procedural posture?
- What did the trial court do? The appellate court below?
- Why is this in the casebook? What subject is it illustrating?

**Holding / rule:**
- What's the holding? One sentence.
- What's the rule that comes out of this case — the portable takeaway?
- How would you phrase the rule if it were in your outline?

**Reasoning:**
- Why did the court decide this way?
- What arguments did the court reject?
- Was there a dissent? What did it argue?

**Application / hypos:**
- What if [fact X] were different — same outcome?
- How does this case compare to [prior case in the syllabus]?
- What's the limiting principle? Where does this rule stop?

**Policy / theory:**
- What's the policy the court is protecting?
- Does this rule make sense? Alternative approaches?

**Professor-specific flavor (from ~/.claude/plugins/config/claude-for-legal/law-student/CLAUDE.md notes):**
- If the professor is known for hypo-heavy calls, weight Application/Hypo questions
- If policy-heavy, weight Policy/Theory
- If fact-heavy socratic (Socratic 101 Paper Chase style), weight Facts + Holding

Pick 6-10 questions across these categories. Rank by likelihood of being asked first (Facts usually go first, then Holding, then the harder categories).

### Step 3: Drill

Use the `socratic-drill` pattern:

1. Ask Question 1. Wait for answer.
2. If right + well-reasoned: acknowledge, move to Question 2.
3. If right but sloppy: don't let it slide. "You got there, but explain — why does the court's reasoning support that?"
4. If wrong: don't give the answer. Ask a narrowing question. "What facts does the court rely on?" Walk them to it.
5. If stuck: narrow further. "Before we go to the holding — what's the procedural posture?"
6. If genuinely lost: tell them to re-read the case. "This is a re-read, not a guess-your-way-through. Come back when you've read it again."

### Step 4: Post-drill summary

At the end:

```markdown
# Cold-Call Prep — [case] — [date]

**Questions drilled:** [N]
**Strong:** [questions where they were confident + right]
**Shaky:** [questions where they guessed or hedged]
**Missed:** [questions where they didn't know]

## Before class tomorrow:
- [specific thing to re-check — facts they got wrong, rule they couldn't state]
- [if shaky on policy/theory: "read the dissent again — that's usually where policy questions come from"]

## Que

Excerpt, first 6,000 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

Install

Pinned to the audited commit cb488a8

Download folder ↗
git clone https://github.com/anthropics/claude-for-legal
git -C claude-for-legal checkout cb488a888d20

The author documents no install command, so this uses the repository directly. Our verdict covers commit cb488a888d20 only.

Works in

The author has not said which agents this skill supports, and we have not tested it in any.

Audit history

Every audit, every source change

  1. ✓SAFE B 89 / 100

    first audit · source cb488a888d20

README badge

For maintainers · shows the latest audit

ModelsAtlas: SAFE BMODELSATLASSAFE · B
[![ModelsAtlas: SAFE B](https://modelsatlas.com/skills/anthropics/cold-call-prep/badge.svg)](https://modelsatlas.com/skills/anthropics/cold-call-prep/)

The badge shows our verdict and grade for this skill, and links back to this page. It changes when the audit does.

Similar skills

Add this badge to your README

[![ModelsAtlas: SAFE B](https://modelsatlas.com/skills/anthropics/cold-call-prep/badge.svg)](https://modelsatlas.com/skills/anthropics/cold-call-prep/)

Questions about cold-call-prep

What does the cold-call-prep skill do?
A suite of plugins for legal workflows
Is cold-call-prep safe to install?
The audit found nothing in the source that contradicts what it says it does, and graded it B (89/100). Grade A is held back for packages that have also passed a sandboxed behavioural run, which is why a clean skill reads B.
What can cold-call-prep access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
How current is this page?
The grade is for one exact copy of the source (cb488a888d20), read on 2026-10-31. The repository is watched, and a new audit runs when it changes — this is the first audit.

Sources

Spotted something wrong? Report a correction · Methodology

Keep going in your AI

Opens with this page as the source. We add nothing else to the prompt.