Skip to main content
skill CX Codex official · Anthropic

cold-start-interview · anthropics, litigation-legal — Claude Code skill: setup and install

The cold-start-interview Claude Code skill builds a litigation practice profile through an interview. Its questions change with the user's role and the side they represent.

Apache-2.0 ★ 10,060 updated 2026-10-23

Install into Claude Code · Plugin marketplace All methods and platforms ↓
/plugin marketplace add anthropics/claude-for-legal
/plugin install litigation-legal@claude-for-legal
Open on GitHub

What cold-start-interview does

The skill interviews the user and saves a practice profile to CLAUDE.md. Its paths differ for in-house counsel, a law firm associate, and a solo practitioner. It also adapts its risk vocabulary to plaintiff work, defense work, or matters on both sides.

The interview captures risk calibration, practice context, and house style. If a profile is already populated, the skill asks before replacing it unless --redo is used. The --check-integrations flag refreshes the MCP connector table without repeating the interview.

The profile is stored at ~/.claude/plugins/config/claude-for-legal/litigation-legal/CLAUDE.md. An integration gets a confirmed mark only after an MCP tool call succeeds; a declaration in .mcp.json alone does not count.

What it helps with

  • create a litigation practice profile for the plugin
  • adapt intake questions to counsel's role and side
  • record risk preferences and document style
  • refresh the list of available MCP integrations

Example prompts for the agent

  • Set up a practice profile for litigation work.
  • Run the interview again and replace the profile with --redo.
  • Check available MCP integrations with --check-integrations.

How to install cold-start-interview

1Plugin marketplace

/plugin marketplace add anthropics/claude-for-legal
/plugin install litigation-legal@claude-for-legal

This installs the whole litigation-legal plugin with all its skills

2Skills CLI (Vercel)

npx skills add anthropics/claude-for-legal --skill cold-start-interview -a claude-code

3Manually: put the files in place

~/.claude/skills/cold-start-interview/SKILL.md

Copy the whole skill folder; for one project use .claude/skills/cold-start-interview/

Step by step, with commands for each platform: How to install skills in Claude Code: five ways

Questions about cold-start-interview

How do I install cold-start-interview in Claude Code?

Add the marketplace with /plugin marketplace add anthropics/claude-for-legal, then install the plugin with /plugin install litigation-legal@claude-for-legal.

What does cold-start-interview do?

It interviews the user about their role, litigation side, risk approach, and document style, then saves those details in a CLAUDE.md practice profile for the litigation plugin.

Can it check integrations without repeating the interview?

Yes. Run /cold-start-interview --check-integrations to rescan MCP connectors and refresh the profile's integrations table.

Source text

SKILL.md · Apache-2.0
---
name: cold-start-interview
description: House cold-start for the litigation plugin — branches by role (in-house, firm associate, solo) and side (plaintiff, defense, both), captures risk calibration, landscape, and house style, and writes the practice profile CLAUDE.md. Use on a fresh install, when the user wants to set up or redo the practice profile, or to re-check available integrations.
argument-hint: "[--redo | --check-integrations]"
---

# /cold-start-interview

1. Check `~/.claude/plugins/config/claude-for-legal/litigation-legal/CLAUDE.md`. If already populated and no `--redo`, ask before overwriting.
2. Follow the workflow and reference below.
3. Run Part 0 (role, side, integration check). The interview branches by role and side.
   - **Role** routes the practice profile structure: **in-house** (portfolio of matters, outside counsel oversight, reserve methodology, board/audit reporting), **firm associate** (case work — matter context, case theory and pivot fact, seed brief in house style, eDiscovery/priv-log setup), or **solo** (caseload + contingency or retainer economics + client expectations + SOL tracking, then the case-theory and brief-style sections).
   - **Side** routes calibration vocabulary: **plaintiff** (asserting, case value, contingency, SOL cliff), **defense** (responding, exposure, reserves where applicable, insurance tender), or **both/varies** (captures a default and lets per-matter skills re-ask).

   After Part 0, walk the sections that match the selected role. Do not run the in-house path for solo users — reserves, ASC 450, and board-memo framing are not the right frame for a solo practice. Offer defaults; capture freeform overrides. Ask for seed documents at each section (non-pushy; note that sharing sharpens every downstream skill).
4. Surface gaps. If the user doesn't have an articulated risk framework or reporting threshold, note it and offer to think through it now or leave `[PLACEHOLDER]` to fill later.
5. Migration: if a populated CLAUDE.md (no `[PLACEHOLDER]` markers) exists at `~/.claude/plugins/cache/claude-for-legal/litigation-legal/*/CLAUDE.md` but not at the config path, copy it to the config path and show the user what was migrated.
6. Write `~/.claude/plugins/config/claude-for-legal/litigation-legal/CLAUDE.md`. Date the footer.
7. Confirm with the user before finalizing: "Here's what I captured — anything wrong?"

## Flags

- `--redo` — re-run the full interview and overwrite `~/.claude/plugins/config/claude-for-legal/litigation-legal/CLAUDE.md`.
- `--check-integrations` — re-scan available MCP connectors and refresh the `## Available integrations` table in `~/.claude/plugins/config/claude-for-legal/litigation-legal/CLAUDE.md` without re-running the full interview. Use after setting up a new connector (DMS, document storage, Gmail, scheduled-tasks, CLM).

When probing: only report ✓ if an MCP tool call actually succeeded. Configured-but-untested connectors should be marked ⚪ with a one-line how-to for confirming. Never report ✓ based on `.mcp.json` declarations alone — that misleads users into thinking something is wired up when it isn't.

---

# Cold-Start Interview: Litigation

## Purpose

Every matter intake, every chronology build, every brief draft, every status rollup reads from this file. If the frame isn't captured, the plugin makes weaker triage calls and the user has to think from scratch each time. This interview fills the frame once so everything downstream gets sharper.

The plugin serves three distinct litigation roles — in-house counsel managing a portfolio of matters, firm associates doing the underlying brief / deposition / discovery work, and solo practitioners running a caseload directly. The vocabulary is different for each, and the interview branches to match. Solo practitioners do not get the in-house path compressed — they get a dedicated solo path (caseload, contingency or retainer economics, client expectations) plus the brief / case-theory sections that apply to anyone who drafts.

The interview also asks which side the user mostly represents — plaintiff (asserting claims), defense (responding to claims), both, or varies by matter. Risk calibration, demand-letter posture, discovery stance, and chronology framing all differ by side, and the practice profile carries the default so downstream skills don't have to ask every time.

**Tone:** socratic, not checklist. If the user doesn't have a written framework, this is often the thing that forces articulation. Lean into that. Don't rush past gaps — name them, offer to think through, allow "leave for later."

## Cold-start check

Read `~/.claude/plugins/config/claude-for-legal/litigation-legal/CLAUDE.md`:
- **Does not exist** → start the interview.
- **Contains `<!-- SETUP PAUSED AT: -->`** → greet the user and offer to resume from that section.
- **Contains `[PLACEHOLDER]` markers but no pause comment** → the template was never completed; offer to start fresh or resume from wherever the placeholders begin.
- **Populated (no placeholders, no pause comment)** → already configured; skip unless `--redo`.

The template structure lives at `${CLAUDE_PLUGIN_ROOT}/CLAUDE.md` — use it as the section scaffold. Write the completed practice profile to the config path, creating parent directories as needed. If a CLAUDE.md exists at the old cache path `~/.claude/plugins/cache/claude-for-legal/litigation-legal/*/CLAUDE.md` but not here, copy it forward.

## Check for the shared company profile

Look for `~/.claude/plugins/config/claude-for-legal/company-profile.md`.

- **If it exists:** Read it. Show a one-line confirmation: "You're [name], [practice setting], at [company], [industry], operating in [jurisdictions]. Right? (Or say 'update' to change the shared profile.)" If confirmed, skip the company questions — go straight to the plugin-specific ones.
- **If it doesn't exist:** You'll be the first plugin this user set up. After the orientation and fork, ask the company questions and write them to the shared profile (per the template at `references/company-profile-template.md` in the plugin root), then continue with the plugin-specific questions. Tell the user: "I've saved your company profile — the other legal plugins will read it and skip these questions."

The company questions that belong in the shared profile (and should NOT be re-asked if it exists): practice setting, company name, industry, what-you-sell, size, jurisdictions, regulators, risk appetite, escalation names. The plugin-specific questions (playbook positions, review framework, house style, supervision model, etc.) stay per-plugin.

## Install scope check

Before the orientation, if you notice the working directory is inside a project (not the user's home directory), flag it. Say once:

> **Heads up — it looks like this plugin may be project-scoped, which means I can only read files in [current directory]. If you'll want me to read documents from elsewhere (Downloads, Documents, Dropbox), install user-scoped instead — see QUICKSTART.md. You can continue with project scope, but you'll need to move files into this folder.**

Ask the user to confirm before proceeding: continue with project scope, or pause to reinstall user-scoped. If the working directory *is* the user's home directory, skip this check silently.

## Before the interview starts

Open with the fork-first preamble. Keep it to 3-4 short lines. Ask quick-or-full before anything else.

> **`litigation-legal` is for people who work litigation — managing a portfolio of matters in-house, drafting briefs and doing discovery at a firm, or both as a solo practitioner.** Not your area? `/legal-builder-hub:related-skills-surfacer`.
>
> **2 minutes** gets you your role (in-house / firm-associate / solo), practice setting, side default (plaintiff / defense), and active matter count, plus working defaults for risk calibration, house brief style, and privilege conventions. **15 minutes** adds your real severity × likelihood bands, settlement-authority ladder (in-house) or fee economics (solo), outside-counsel roster, house brief style from a seed brief, privilege-log format, demand-letter templates, and landscape notes.
>
> Quick or full? (Upgrade any time with `/cold-start-interview --full`.)

**Quick start path:** ask only Part 0 (role, practice setting, integrations) and the path branch. Write the config with `[DEFAULT]` markers on everything else. Close with: "Done. You can start using the commands now. I've used sensible defaults for risk calibration, house style, and case-theory scaffolding. When a skill's output feels off, that's usually a default you should tune — it'll tell you which. Run `/litigation-legal:cold-start-interview --full` anytime to do the whole interview, or `/litigation-legal:cold-start-interview --redo <section>` to re-do one part."

**Full setup path:** the existing interview flow below. After the user picks, give the fuller orientation described next, then proceed to Part 0.

## After the user picks quick or full

Give the fuller orientation. One paragraph, in your own voice:

> "This plugin maintains: your practice profile (risk calibration, privilege conventions, house style), a matter ledger (`_log.yaml`), per-matter files (chronology, hold notices, histories, priv logs), and a work-product archive. It supports litigation work whether you're in-house managing a portfolio, a firm associate drafting briefs and depo outlines, or a solo practitioner doing both. It learns which role you're in, your risk calibration or case theory, your dispute landscape or production setup, your house conventions, and writes them into a plain-text file the plugin reads from every time. Everything you answer can be changed later."

Then the fresh-profile note:

> "Setup builds a fresh professional profile from your answers. It does not read your personal Claude history, other conversations, or your home-directory CLAUDE.md. If I notice relevant information in our conversation context — e.g., you mentioned your company or matter earlier — I'll ask before using it. Nothing personal gets folded into your practice configuration unless you type it or approve it."

Then: "Ready? A few quick questions first."

**Why this matters** (offer if the user pushes back on the time cost). Every matter intake, every portfolio status, every brief draft reads from the configuration this interview writes. A generic configuration gives generic output — a default risk matrix, a default citation style, a generic priv-log format. Telling the plugin the actual severity bands, the actual settlement authority ladder, the actual brief structure is what makes the difference between "a litigation AI tool" and "a tool that triages and drafts the way you do." Especially load-bearing: the pivot fact (if firm-side) and the seed documents.

Draw the practice profile only from the user's typed answers and documents they upload during the interview. Do not read `~/CLAUDE.md` or pull practice facts from ambient context. If something relevant is already visible in this conversation, ask before using it.

## Interview pacing

- **Assume the answer exists somewhere.** When a question asks for information that's probably written down somewhere — company description, playbook, escalation matrix, style guide, handbook, jurisdiction list, matter portfolio — prompt for a link or a paste before asking the user to type it from memory. "Paste a link or a doc, or give me the short version" is the default ask for anything that's more than a sentence. An interviewer who makes people re-type what they've already written has failed the first job of an interviewer.

**Pause for real answers.** Some questions have quick tap-through answers. Others need the user to type something, describe something, or upload an exemplar (board memo, hold template, demand letter, risk memo, case theory memo, seed brief). When a question needs more than a quick tap:

- **Batch size — count subparts.** "Never ask more than 2-3 questions in one turn" means 2-3 *answerable prompts*, counting subparts. One question with 5 subparts is 5 questions. The test: can the user answer without scrolling? If the questions don't fit on one screen, it's too many. Prefer structured tap-through questions where possible — they don't require scrolling or typing.
- **Ask the question and wait.** Say explicitly: "This one needs a typed answer — I'll wait." Do not move to the next question until the user responds. This matters most for the theory section (firm-associate path) — do not paraphrase a half-answer and push on.
- **For seed-document uploads:** "Paste the contents, share a file path, or say 'skip for now.' If you skip, I'll flag the gap in your practice profile so you can fill it later." Then actually wait.
- **Before writing the practice profile:** review every captured answer. List any questions that were skipped, answered with placeholders, or produced a contradiction. Say: "Before I write your practice profile, here's what's still open: [list]. Want to fill any of these now, or leave them as placeholders?" Then wait.
- **Never** write a practice profile with silent gaps. Every `[PLACEHOLDER]` should be a deliberate choice the user made to skip, not a question that scrolled past. The `LIMITED DATA` footer is for seed-document thinness only — not for questions the interview never actually asked.
- **Pause and resume.** Tell the user up front: "If you need to stop, say 'pause' (or 'stop', or 'let me come back to this') and I'll save your progress. Run `/litigation-legal:cold-start-interview` again later and I'll pick up where you left off." When the user pauses, write a partial configuration with a `<!-- SETUP PAUSED AT: [section name] — run /litigation-legal:cold-start-interview to resume -->` comment at the top and `[PENDING]` markers (distinct from `[PLACEHOLDER]`) on unanswered fields. When setup re-runs and finds a paused config, greet: "Welcome back. You paused at [section]. Your earlier answers are saved. Pick up where we left off, or start over?" Do not re-ask questions already answered.

**Verify user-stated legal facts as they come up in setup.** When the user answers an interview question with a specific rule citation, statute number, case name, deadline, threshold, jurisdiction, or registration number — and it's something you can sanity-check — do the check before writing it into the configuration. If what they said conflicts with your understanding or with something they've pasted, surface it: "You said the threshold is X; my understanding is Y — can you confirm which goes in the profile? `[premise flagged — verify]`" A wrong fact written into CLAUDE.md propagates into every future output; catching it here is one of the highest-leverage moments in the product.

## Part 0: Who's using this + role routing

### Who's using this?

> Who'll be using this plugin day to day? (This feeds the work-product header on every matter briefing, chronology, priv log, and demand draft — lawyer outputs get the privilege header, non-lawyer outputs get the "research notes, review with counsel" header.)
>
> 1. **Lawyer or legal professional** — attorney, paralegal, legal ops working under attorney oversight.
> 2. **Non-lawyer with attorney access** — founder, business lead, contracts manager, HR, procurement; you have an in-house or outside attorney you can consult.
> 3. **Non-lawyer without regular attorney access** — you're handling this yourself.

If the answer is 2 or 3, say this once (don't repeat it on every output):

> You can use every feature here — research, review, drafting, tracking. Two things change in how I work:
>
> 1. **I'll frame outputs as research for attorney review, not as verdicts.** Instead of "GREEN — sign it," you'll get "here's what I found and here are the questions to ask before you sign." That's more useful than a green light you can't be sure of.
> 2. **I'll pause before steps that have legal consequences** — sending a demand, responding to a subpoena, issuing or releasing a legal hold, filing a brief, submitting a privilege log, designating documents in discovery, closing a matter, accepting a settlement. I'll ask whether you've reviewed with an attorney, and I'll put together a short brief so the conversation with them is fast.
>
> This isn't a disclaimer. It's the plugin knowing the difference between what it's good at — research, organization, structure — and licensed legal judgment about your specific situation, which a tool can't give you. A few hours of a lawyer's time at the right moment is usually cheaper than the mistake.

If the answer is 3, add:

> If you need to find a licensed attorney, solicitor, barrister, or other authorised legal professional in your jurisdiction: your professional regulator's referral service is the fastest starting point (state bar in the US, SRA/Bar Standards Board in England & Wales, Law Society in Scotland/NI/Ireland/Canada/Australia, or your jurisdiction's equivalent). Many offer free or low-cost initial consultations.

### Role (the branching question — ask early)

> **How do you work litigation?** (This determines which pillars of the interview run — in-house gets reserves and board memos, firm-associate gets case theory and seed briefs, solo gets caseload economics plus the firm-associate brief work. It also sets defaults for /matter-intake, /portfolio-status, /oc-status, and every other skill's vocabulary.)
>
> **(a) In-house managing a portfolio** — matters, outside counsel, deadlines, demands, holds. You own many matters at once, most of which are run by outside firms. Status rollups and board memos are part of your job.
>
> **(b) At a firm doing brief drafting, discovery, deposition prep, document review** — you're the associate or paralegal responsible for actually producing the work product. One or a few matters, deep on each.
>
> **(c) Solo / small firm running a caseload** — you intake, triage, advise, and draft. No partner above you; no in-house reserve / board-memo layer. Economics are contingency or retainer, not billable hours to a large client.
>
> **(d) Something else** — describe in a sentence.

Record the answer in the practice profile's `## Role` section at the top (`in-house | firm-associate | solo | other`). Downstream skills read this to pick defaults (e.g., chronology mode, which commands are primary, which vocabulary to use).

**Branching rules for the rest of this interview:**

- `in-house` → run the **In-house path** (Pillars 1–3 below). Skip the firm-associate and solo sections.
- `firm-associate` → run the **Firm-associate path** (Parts A–D below). Skip the in-house portfolio / OC / board-memo questions and the solo caseload / economics questions.
- `solo` → run the dedicated **Solo path** (Sections S1–S3 below) — caseload, client expectations, contingency or retainer economics, office management — **then** run the Firm-associate path (Parts A–D) because solo practitioners still write briefs and work cases. Do NOT run the In-house path — reserves, ASC 450, board memos, and settlement-authority ladders up to a GC are not the right frame for a solo practice.
- `other` → ask for a one-sentence description, then pick the closest branch.

### Which side do you mostly represent?

Ask this right after the role question. It's load-bearing for risk-calibration framing, demand-letter posture, discovery stance, and the way chronologies are built.

> **Which side do you mostly represent?** (This feeds /demand-draft, /demand-received, /subpoena-triage, /chronology, and /claim-chart — plaintiff framing treats demand letters as assertions and discovery as offensive, defense framing treats them as received and responsive.)
>
> **(a) Plaintiff / claimant** — you bring claims for individuals or businesses. Demand letters are assertions you draft and send. Discovery is offensive. Statute of limitations is a cliff you work against. Economics are often contingency.
>
> **(b) Defense / respondent** — you defend businesses or individuals against claims. Demand letters are received and triaged. Discovery is defensive. Exposure is assessed, reserved (in-house), tendered to insurance (where applicable).
>
> **(c) Both** — your practice regularly includes both. Ask for a default (plaintiff or defense); individual skills will ask per-matter when it matters.
>
> **(d) Varies by matter** — no strong default; every matter gets asked.

Record under `## Side` in the practice profile (`plaintiff | defense | both [default plaintiff/defense] | varies`). Branching rules for calibration that follows:

- **Plaintiff:** risk calibration is about case value, contingency economics, client expectations, statute of limitations exposure. Demand letters are the assertion. Discovery is offensive. Settlement-authority conversations are with the client, not a GC/board. (For firm-associate plaintiff-side: partner review replaces GC escalation.)
- **Defense:** risk calibration is about exposure, reserves (in-house only), settlement authority, insurance coverage. Demand letters are received and triaged. Discovery is defensive — responding, asserting privilege, narrowing.
- **Both / varies:** the interview captures the default and the skills (`demand-draft`, `subpoena-triage`, `matter-intake`, `chronology`, `claim-chart`) ask per-matter when the side changes the output.

### Practice setting

> Which best describes where you're practicing?
>
> 1. **Solo practitioner**
> 2. **Small firm (2–10)**
> 3. **Midsize firm**
> 4. **Large firm / Am Law**
> 5. **In-house** (company legal department)
> 6. **Government**
> 7. **Legal aid**
> 8. **Clinic**
> 9. **Other**

This refines escalation / supervision language in the practice profile:

- **Solo / small without hierarchy (1, 2):** Reframe authority-ladder questions as "when do you call in outside counsel or a colleague for a second opinion." Escalation maps to *consult* not *route for approval*.
- **Midsize / large firm / in-house / government (3, 4, 5, 6):** Ask the full escalation chain, authority ladder, and internal-contacts table.
- **Legal aid / clinic (7, 8):** Route toward the supervision model — supervising attorney of record, sign-off chain, review-queue mechanics.
- **Other (9):** Ask for a one-sentence description, then pick the closest branch.

**Practices that don't fit the boxes.** If the user's practice doesn't match the options above (international arbitration, public international law, amicus-only, academic consulting, pro bono panel, tribal court, military justice, maritime, or anything else the standard categories assume away), offer: "It sounds like your practice doesn't fit my usual categories. Tell me about it in your own words — what you do, who for, what jurisdictions and forums, what the work looks like — and I'll build your profile from that instead of forcing you into boxes that don't fit. I'll skip or adapt the questions that don't apply." Then build the profile from the free-form description, flagging which template fields were filled, adapted, or left empty because they don't apply. A profile built from a forced fit is worse than a sparse profile built from what's actually true.

### What's connected?

> This plugin can work with: DMS (iManage), document storage (Google Drive, SharePoint, Box), Gmail, scheduled-tasks, CLM (Ironclad), eDiscovery (Everlaw, Relativity, DISCO, Aurora), legal research (CourtListener, Descrybe, Trellis), outside-counsel recommendations (TopCounsel). Let me check which connectors you have configured — features that need them will work, and features that don't will fall back gracefully instead of failing silently.

**Check what's actually connected, not what's configured.** A connector listed in `.mcp.json` is *available*. A connector that's actually responding is *connected*. These are different, and confusing them destroys trust. For each connector this plugin uses:

- If you can test the connection (call a simple MCP tool like a list or search), report ✓ only on a successful response.
- If you can't test (no way to probe from here), report ⚪ "configured but not verified — open your MCP settings to confirm" with a one-line how-to.
- Never report ✓ based on configuration alone.

For connectors that show as not connected, tell the user how to connect. Example phrasing: "Box isn't connected. In Claude Cowork: Settings → Connectors → Add → Box → sign in. In Claude Code: add the Box MCP to your config or via `/mcp`. This plugin works without it — you'll paste documents instead of pulling them — but connecting it makes document pulls automatic."

Then report findings in this form:

> - ✓ [Integration] — connected (tested)
> - ⚪ [Integration] — configured but not verified. Open your MCP settings to confirm.
> - ✗ [Integration] — not found. [Feature] will fall back to [manual alternative]. [How to connect.]

You don't need all of these. Core features work with file access alone.

Write a `## Role`, `## Who's using this`, and `## Available integrations` section into the plugin config immediately after the opening. Add `## Outputs` with the work-product header rule per the CLAUDE.md template.

---

## In-house path (role == `in-house`)

*Skip this whole section if the user's role is `firm-associate` or `solo`.*

> I want to capture the frame you triage matters against — your risk calibration, the dispute landscape, and how you write. Once, so every matter intake reads from it. I'll offer defaults where there are reasonable ones. You can accept, edit, or leave blank to come back to.
>
> I'll also ask for seed documents along the way — prior board memos, reserve memos, litigation hold templates, exemplar demand letters, a sample risk memo. Ten to twenty total across the interview is the target. Anything below ten and I'll flag the practice profile as LIMITED DATA in the footer — skills will still run, but their outputs will be thinner because they're matching on weaker patterns. Templates-first: if you upload an exemplar, I'll read it and only ask about gaps rather than walking the full structure from scratch.

### Pillar 0 — Company profile

Team-level context. If another `-legal` plugin already has a `## Company profile` block populated, copy it here rather than re-enter.

- Org / legal entity
- Industry
- Public / private / subsidiary
- Regulated status
- Core jurisdictions (operational + frequent-fora)
- Headcount + legal team size
- Key internal contacts (GC, CFO, HR lead, Comms, CISO, Board lit/audit chair) — names + when to loop in
- This counsel's name and reporting line

### Pillar 1 — Risk calibration

> Before the structured questions: do you have an existing risk-calibration memo, a reserve-policy document, or an outside-counsel billing-guidelines doc I can read? Paste the contents, share file paths, or say 'no' and I'll walk the pillar question by question. If you share one, I'll extract the severity bands, materiality thresholds, and authority ladder and only ask about gaps.

If not:

**Risk appetite (2 min)** — in a sentence, how does this company approach litigation? (This feeds /matter-briefing and /portfolio-status — sets how conservative or aggressive every matter briefing is when calling a matter's risk tier.)

**Severity × likelihood (3–5 min)** — offer the default 3×3. Severity bands (dollar and non-dollar triggers). Likelihood bands. If unarticulated: "Fair. A lot of counsel don't. Want to sketch now, or leave the default?"

**Materiality thresholds (2–3 min)** — reserve trigger, disclosure trigger, board/audit committee, GC-only escalation. *Seed doc opportunity:* reserve memo template or disclosure checklist.

**Settlement authority (1–2 min)** — dollar ladder, special carve-outs (structural relief requires board regardless of dollar).

**Plain-English escalation (1 min).** Ask directly:

> When a matter needs something above your authority — a settlement offer above your band, a demand you can't answer alone, a hold decision that needs the GC — who does that go to? Give me a name, a role, or "I decide myself."

(Solo practitioners: "I decide myself" is the right answer; the question still matters for the record. If you loop in outside counsel for second opinions, name the firm.)

**Insurance profile (1–2 min)** — lines in force (D&O, EPL, Cyber, GL/E&O), carriers, limits, retentions, tendering protocol.

**Offer:** "If you didn't upload a risk-calibration memo, want me to write your risk calibration and authority ladder up as a standalone memo you can share and maintain?"

### Pillar 2 — Landscape

*Company profile lives in Pillar 0. Landscape is litigation-specific.*

- Business context (30 sec) — one-paragraph on what we do and why we get sued.
- Dispute patterns (2–3 min) — matter types, frequency, posture.
- Frequent adversaries (1–2 min).
- Outside counsel bench (2–3 min) — firms, lead partners, matter type, rate posture, engagement letter status. *Seed doc:* outside counsel guidelines. (This feeds /oc-status — the skill later drafts weekly status requests to these firms.)
- Frequent fora (30 sec).
- Document storage (2–3 min) — where matter docs live (filesystem, Drive, SharePoint, Box, Gmail, CLM, DMS, eDiscovery), default matter folder pattern, how docs get shared with OC.
- Conflicts clearance (1–2 min) — how this shop runs conflicts; who does it; hard block on intake or parallel.

### Pillar 3 — House style

> Before the structured questions: do you have a house-style guide, a template board memo, a hold-notice template, or exemplar demand letters I can read? Paste the contents, share file paths, or say 'no' and I'll walk the questions.

If not:

- Board / audit committee memo (2 min) — format, tone, cadence. *Seed doc:* recent board memo (redacted fine).
- Reserve memo — format and approver. *Seed doc:* sample reserve memo.
- Outside counsel directives — email format, cadence, budget posture.
- Privilege conventions — marking; default subjective-call posture (mark and flag); review mechanic (inline / queue / both)...

© Anthropic · Apache-2.0 · anthropics/claude-for-legal

Similar skills

All skills for Agent setup: memory, context, orchestration →

cold-start-interview by other authors comes first.

cold-start-interview official anthropics cold-start-interview runs an interview about your AI governance practice and writes the profile to CLAUDE.md so the plugin's other skills read the same setup. CX Codex +1 673installs cold-start-interview official Anthropic cold-start-interview — a Claude Code skill that configures a regulatory watchlist, indexes policy files, and records a materiality threshold. Run it for initial setup, a reset, or an integration check. CX Codex +1 ★ 10,060repo stars cold-start-interview official Anthropic The cold-start-interview Claude Code skill builds a product counsel profile from past launch reviews and connected services. Run it for initial setup, counsel onboarding, or a fresh risk calibration. CX Codex +1 ★ 10,060repo stars cold-start-interview official Anthropic The cold-start-interview Claude Code skill interviews a team about its privacy practice and saves the profile to CLAUDE.md. It starts during initial setup, when that file needs completing, or when integrations need another check. CX Codex +1 ★ 10,060repo stars cold-start-interview official Anthropic The cold-start-interview Claude Code skill records a law school clinic’s practice areas, jurisdiction, and supervision model. Run it for initial setup, a fresh configuration, or an integration check. CX Codex +1 ★ 10,060repo stars court-rules Court Rules court-rules is a Claude Code and Codex plugin that searches US federal and state filing rules, judges’ standing orders, and court holidays. Results cite the court’s own documents. CX Codex ★ 1,596repo stars cold-start-problem wondelai cold-start-problem — a skill for Claude Code, Cursor, Codex and other agents that applies Andrew Chen's network-effects framework to product launches. CX Codex +1 3,400installs cold-start-strategy kostja94 cold-start-strategy guides cold start planning for AI and SaaS products, helping find seed users and launch when starting with zero traction. CX Codex +1 1,200installs cold-start vivy-yi The cold-start skill lays out a launch plan for a new Xiaohongshu account, from preparing posts to testing formats and using audience response to choose a direction. CX Codex +1 175installs discovery-interview parcadei discovery-interview runs a detailed product interview that turns a vague idea into a specification. Use it when requirements need to be worked out before implementation. CX Codex +1 4,200installs core-interview pwguler core-interview — a skill for Claude Code that runs an interview loop through decision trees until every branch is settled. CX Codex +1 1,000installs

From the same set174

The whole set →
customize customize — a skill for Claude Code edits a single setting in a privacy practice profile. It responds to requests to update the profile, a playbook, or a specific section. CX Codex +1 ★ 10,060repo stars cold-start-interview The cold-start-interview Claude Code skill interviews a team about its privacy practice and saves the profile to CLAUDE.md. It starts during initial setup, when that file needs completing, or when integrations need another check. CX Codex +1 ★ 10,060repo stars matter-workspace The matter-workspace Claude Code skill manages separate workspaces for legal matters. It creates, lists, switches, and archives matters, or detaches the active one. CX Codex +1 ★ 10,060repo stars customize The customize Claude Code skill edits individual settings in a litigation practice profile. It runs when someone asks to update the profile or change a specific setting. CX Codex +1 ★ 10,060repo stars docket-watcher docket-watcher — a subagent for Claude Code, Codex, and Cursor. Sweeps court dockets on a schedule, pulls filings, computes candidate deadlines, and writes status reports. CX Codex ★ 10,060repo stars customize The customize Claude Code skill updates a legal clinic profile on request, from its jurisdiction to practice templates and semester settings. It activates for requests to revise the config, add a practice area, or start a new semester. CX Codex +1 ★ 10,060repo stars cold-start-interview The cold-start-interview Claude Code skill records a law school clinic’s practice areas, jurisdiction, and supervision model. Run it for initial setup, a fresh configuration, or an integration check. CX Codex +1 ★ 10,060repo stars customize customize — a Claude Code skill for editing a Legal Builder Hub profile. It responds to requests to change profile details, registries, update settings, or QA strictness. CX Codex +1 ★ 10,060repo stars

Project sponsors

Companies that keep this analytics open