Intake a new matter — uniform questions covering identification, conflicts, source, risk triage, materiality, outside counsel, owners, legal hold, and key dates; writes matter.md and history.md and appends a structured row to _log.yaml. Use when the user says "new matter", "intake this matter", or wants to bring a new matter into the portfolio.
Standardmäßig ist der Prompt ausgewählt, der zuerst die Quelle prüft. Sie können zu einem direkten Befehl wechseln oder eine lokale Kopie herunterladen.
Quelldateien prüfen
Lesen Sie SKILL.md und alle von SkillsMP angezeigten Begleitdateien, bevor Sie sich für eine Installation entscheiden.
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fügen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prüfen und installieren.
Ein direkter Befehl überspringt den Prüf-Prompt. Prüfen Sie die Quelle, bevor Sie ihn ausführen.
Intake a new matter — uniform questions covering identification, conflicts, source, risk triage, materiality, outside counsel, owners, legal hold, and key dates; writes matter.md and history.md and appends a structured row to _log.yaml. Use when the user says "new matter", "intake this matter", or wants to bring a new matter into the portfolio.
Codex v1 local-input note: This migrated skill supports local files and pasted text by default. References to Drive, CLM IDs, Slack, Westlaw, iManage, Ironclad, eDiscovery, dockets, or other remote systems require a separately configured Codex connector/MCP server. When a connector is unavailable, ask for a local export, local file path, or pasted excerpts. If config/local/codex-for-legal/<practice>/CLAUDE.md is missing, ask the user to run the relevant cold-start-interview or customize skill and copy from config/templates/codex-for-legal/<practice>/CLAUDE.md.
$litigation-legal:matter-intake
Load config/local/codex-for-legal/litigation-legal/CLAUDE.md → risk calibration (for triage), landscape (for context, conflicts method), stakeholders (for who to loop in).
Generate slug from matter name (lowercase, hyphens, year).
Create config/local/codex-for-legal/litigation-legal/matters/[slug]/matter.md — full narrative intake.
Create config/local/codex-for-legal/litigation-legal/matters/[slug]/history.md — seeded with the intake as the first entry.
Append structured row to config/local/codex-for-legal/litigation-legal/matters/_log.yaml.
Confirm with the user: "Here's the row I'll write — any edits?"
Matter Intake
Purpose
Every new matter goes through the same intake so the portfolio stays comparable. Uniform rows in _log.yaml let the status skill roll up. Narrative in matter.md captures what the row can't. History file seeded here becomes the event record.
If the practice profile's ## Side is plaintiff, defense, or a "both — default X" variant, pre-fill the role from that default and confirm. If ## Side is varies by matter, ask cold. Never silently assume a posture the practice profile hasn't set.
The role drives downstream skills: plaintiff-posture matters route risk triage to case value / contingency economics; defense-posture matters route to exposure / reserves / insurance tender.
Jurisdiction (court, arbitration forum, or regulatory body)
2. Conflicts check
Before going further, run the conflicts step per config/local/codex-for-legal/litigation-legal/CLAUDE.md → Conflicts clearance.
Status:cleared | pending | not-run | waived
Method: match what config/local/codex-for-legal/litigation-legal/CLAUDE.md declares (corporate-legal | outside-counsel | system-check | informal | other). If the declared method is informal, say so — the record still captures that a counsel's-judgment check was the basis.
Cleared by: name / team / firm
Cleared date: YYYY-MM-DD
Checked against: brief list of the specific names/entities run (counterparty, known affiliates, adverse counsel if known, key witnesses). Thin is fine; "no" is not.
Notes: anything flagged but cleared (e.g., "Smith on our board sat on counterparty's board 2019–2021 — cleared as non-overlapping to this matter").
Behavior by status:
cleared → proceed.
pending → proceed with intake; flag prominently in matter.md and in the log row that conflicts are outstanding; surface again on every $litigation-legal:matter-update and in $litigation-legal:portfolio-status until resolved.
waived → rare; requires a conflict-waiver rationale (writing the waiver is outside this skill — capture that one exists, who signed it, and where it lives).
not-run → STOP. This is a gate. The skill will not create matter.md, history.md, or a _log.yaml entry until the conflicts posture is resolved. Three acceptable paths:
Path 1 — Run conflicts now. Pause this intake. Clear per config/local/codex-for-legal/litigation-legal/CLAUDE.md Conflicts clearance. Return with status: cleared or status: waived with rationale.
Path 2 — Mark pending with owner + due date. Allowed only when config/local/codex-for-legal/litigation-legal/CLAUDE.md Conflicts clearance declares parallel-intake acceptable. Capture: who is running conflicts, when they're expected to return, what entities they're checking. Intake proceeds; matter row carries conflicts.status: pending; $litigation-legal:portfolio-status flags it every run; $litigation-legal:matter-update re-prompts until resolved.
Path 3 — Bypass with documented rationale. Only if the user explicitly acknowledges the bypass. Record in conflicts.override:
This field is visible in every $litigation-legal:portfolio-status, every /matter briefing, and every $litigation-legal:matter-update until removed. It is never removed by the skill — only by explicit user edit to _log.yaml after conflicts are actually cleared.
Do not proceed silently. "I'll do it later" is not an acceptable response. One of Path 1/2/3 must be chosen, and the choice is captured in the record.
This step is not about the skill deciding whether a conflict exists — that's the user's/firm's judgment. It's about making sure the check happened and the record reflects it.
If the risk calibration in config/local/codex-for-legal/litigation-legal/CLAUDE.md is thin, don't fake precision. Use the user's gut and note the thinness.
5. Materiality
Against the house thresholds in config/local/codex-for-legal/litigation-legal/CLAUDE.md:
reserved | disclosed | monitored | none
If reserved: reserve amount and whether finance has been notified
If disclosed: filing and footnote location
6. Outside counsel
Firm
Lead partner
Lead partner email (used by $litigation-legal:oc-status to draft status requests)
Engagement letter status: signed | pending | none
Budget authorization: amount and approver
Seed doc opportunity: "Engagement letter path, if signed."
If risk is medium or higher and no outside counsel is assigned — flag it.
7. Internal owners
From config/local/codex-for-legal/litigation-legal/CLAUDE.md landscape — which internal stakeholders are involved?
Business lead
HR partner (if employment)
Comms contact (if reputational risk)
CISO (if data or cyber)
Other
8. Legal hold
Issued? If yes: date, scope, custodians (list of names).
Next refresh date (default: six months from issuance; adjust per matter).
If no and this is active litigation or reasonably anticipated: flag urgently; offer to run $litigation-legal:legal-hold [slug] --issue after intake completes.
Seed the history file with the intake as entry zero:
# History: [Matter Name]
Append-only event log. Most recent at top.
---
## [YYYY-MM-DD] — Matter opened
[Source, who brought it in, initial triage summary, outside counsel assigned, legal hold issued yes/no.]
Append to config/local/codex-for-legal/litigation-legal/matters/_log.yaml
Add a row per the schema. Example:
-id:acme-v-us-2026name:"Acme Corp v. Company"type:contractrole:defendantcounterparty:"Acme Corp"jurisdiction:"N.D. Cal."# status is derived from source:# source: pre-suit-threat | demand-letter → status: threatened# source: complaint-served | subpoena | regulator-inquiry → status: active# source: internal-report → status: threatened (default) or active if formal process has startedstatus:activestage:pleadingssource:complaint-servedoutside_counsel:firm:"Wilson Sonsini"lead:"J. Reyes"email:"jreyes@wsgr.example.com"engagement:signedconflicts:status:clearedmethod:corporate-legalcleared_by:"K. Patel"cleared_date:2026-04-20override:# populated only on Path 3 bypassby:nulldate:nullrationale:nullrisk:highmateriality:reservedexposure_range:"$2M–$5M"internal_owners:business_lead:"Jane Smith"hr_partner:nullcomms_contact:nulllegal_hold:issued:trueissued_date:2026-02-15scope:"Sales org 2023–2026"custodians: ["Jane Smith", "R. Chen", "T. Patel"]
last_refresh:2026-02-15next_refresh:2026-08-15released:nullrelated_matters: []
opened:2026-04-20next_deadline:2026-05-15last_updated:2026-04-20path:matters/acme-v-us-2026/
Confirm before writing
Show the user the row and the matter.md content:
Here's what I'll write. Flag anything wrong or thin before I commit.
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
Run the conflicts check itself. It records the result, status, method, and the entities checked. The actual clearance happens in whatever system (or judgment) the house practice profile declares. If the user says "cleared," the skill takes that at face value and captures the metadata.
Decide the initial theory. It captures what the user says; it doesn't invent one.
Issue the legal hold. Flags it if missing. User issues it.
