Communitygithub.com

ThomasMoreAI/legal-skills-open

Analyzes an entire portfolio of contracts at scale — extracting key terms, obligations, renewal dates, risk provisions, and financial commitments across hundreds or thousands of agreements simultaneously. Produces structured reports, obligation trackers, renewal calendars, and risk matrices. Also supports playbook-based redlining where contracts are compared against preferred positions and automatically marked up. Use when an in-house legal team, GC, legal ops, procurement, or finance team needs portfolio-wide visibility into contract terms, upcoming renewals, obligation compliance, or batch contract review against organizational standards.

legal-skills-open 是什么?

legal-skills-open is a Claude Code agent skill that analyzes an entire portfolio of contracts at scale — extracting key terms, obligations, renewal dates, risk provisions, and financial commitments across hundreds or thousands of agreements simultaneously. Produces structured reports, obligation trackers, renewal calendars, and risk matrices. Also supports playbook-based redlining where contracts are compared against preferred positions and automatically marked up. Use when an in-house legal team, GC, legal ops, procurement, or finance team needs portfolio-wide visibility into contract terms, upcoming renewals, obligation compliance, or batch contract review against organizational standards.

兼容平台~Claude Code~Codex CLI~Cursor
npx skills add https://github.com/ThomasMoreAI/legal-skills-open/tree/HEAD/general/contracts/skills/contract-portfolio-intelligence

在你喜欢的 AI 中提问

打开一个已预加载此 Agent Skill 的新对话。

文档

Contract Portfolio Intelligence

Performs portfolio-scale contract analysis: batch extraction of terms, obligations, and risks across a large set of agreements, plus playbook-based redlining against organizational standards.

Prerequisites

Before starting, gather:

  1. Contract corpus — uploaded documents (PDF, DOCX) or a vault/folder reference containing the agreements
  2. Analysis scope — what to extract (renewals, indemnities, payment terms, liability caps, termination rights, or custom fields)
  3. Playbook (if redlining) — the organization's preferred positions, fallback clauses, and non-negotiable terms
  4. Output format — table/spreadsheet, narrative report, or both
  5. Audience — who receives the output (GC, board, finance, procurement)

Workflow

Phase 1: Ingest and Classify

  1. Process each contract through OCR if scanned; extract full text from native PDFs and DOCX files
  2. Identify contract type (NDA, MSA, SaaS agreement, employment, vendor, lease, SOW, amendment, etc.)
  3. Detect parent-child relationships (master agreements → amendments → SOWs → change orders)
  4. De-duplicate — flag identical or near-identical documents
  5. Tag metadata: parties, effective date, governing law, contract type

Output a contract inventory table:

#DocumentTypePartiesEffective DateGoverning LawParent Doc
1ExampleMSAA ↔ B2024-01-15Delaware—

Phase 2: Term Extraction

For each contract, extract the following structured fields (skip fields that don't apply to the contract type):

Dates & Term

  • Effective date, expiration date, term length
  • Auto-renewal (yes/no), renewal notice period, notice deadline
  • Termination for convenience (yes/no), notice period, cure period

Financial Terms

  • Contract value / total commitment
  • Payment terms (Net 30, Net 60, etc.)
  • Price escalation / adjustment mechanisms
  • Minimum commitments or volume thresholds
  • Late payment penalties

Risk Provisions

  • Indemnification — mutual vs. one-sided, carve-outs, cap
  • Limitation of liability — aggregate cap, per-incident cap, excluded damages
  • Insurance requirements — types, minimum coverage amounts
  • Warranty — scope, duration, remedy limitations
  • Governing law and dispute resolution (litigation vs. arbitration, venue)

Obligations

  • Deliverables and milestones with deadlines
  • Reporting or audit rights
  • Data protection / security obligations
  • Non-compete or exclusivity restrictions
  • Confidentiality term and surviving obligations

IP & Data

  • IP ownership and license-back provisions
  • Data handling, return, and destruction requirements
  • Usage rights post-termination

Store extracted data in a structured format (one row per contract, one column per field).

Phase 3: Portfolio Reports

Generate the reports requested by the user. Common report types:

Renewal Calendar

  • All contracts with renewal/expiration dates within a specified window
  • Sorted by urgency (nearest deadline first)
  • Flag auto-renewals where notice deadline is approaching
  • Include counterparty, contract value, and required action

Risk Matrix

  • Each contract scored on risk dimensions: indemnification exposure, liability cap adequacy, insurance gaps, unfavorable termination terms
  • Risk level per dimension: High / Medium / Low
  • Aggregate portfolio risk summary
  • Prioritized list of contracts requiring renegotiation

Obligation Tracker

  • Active obligations by contract with deadlines
  • Grouped by responsible team (legal, finance, ops, IT)
  • Status: upcoming / due / overdue / completed
  • Escalation flags for overdue items

Financial Summary

  • Total portfolio commitment by period (quarterly/annual)
  • Breakdown by contract type, counterparty, or business unit
  • Contracts approaching minimum commitment thresholds
  • Payment term distribution

Custom Extraction

  • User specifies any data point via natural language prompt
  • Agent extracts that field across all contracts
  • Returns structured table with contract reference, extracted value, and confidence

Phase 4: Playbook Redlining

When the user provides a playbook of preferred positions:

  1. Parse the playbook — extract each preferred clause, its fallback positions, and non-negotiable terms
  2. Compare each contract against the playbook provision-by-provision:
    • ✅ Matches preferred position → no action
    • ⚠️ Deviates but within fallback range → flag with suggested language
    • ❌ Violates non-negotiable → flag as critical, provide replacement language
  3. Generate a redline summary per contract:
    • Section reference, current language, preferred language, risk level, recommended action
  4. Produce a portfolio deviation report:
    • Which contracts deviate most from standards
    • Most common deviations across the portfolio
    • Priority queue for renegotiation based on risk × contract value

Playbook Structure

A playbook should define positions for key provisions. Example structure:

Provision: Limitation of Liability
  Preferred: Mutual cap at 12 months of fees; carve-outs for IP infringement and confidentiality breach
  Fallback: Mutual cap at 24 months of fees
  Non-negotiable: Must have mutual cap; unlimited liability is rejected

Provision: Indemnification
  Preferred: Mutual indemnification for third-party IP claims and confidentiality breach
  Fallback: One-sided indemnification acceptable if capped
  Non-negotiable: Must include IP infringement indemnity

Provision: Termination for Convenience
  Preferred: Either party, 30 days written notice
  Fallback: 60-90 days notice acceptable
  Non-negotiable: Must have termination for convenience right

Quality Checks

After generating any output:

  • Every contract in the corpus appears in the inventory — none skipped
  • Extracted dates are validated (no future effective dates on executed contracts, expiration after effective)
  • Parent-child relationships are logically consistent
  • Risk scores are justified with specific clause references
  • Renewal deadlines account for notice periods (deadline = expiration minus notice period)
  • Financial figures are in consistent currency and units
  • Playbook deviations cite the specific section and language from both the contract and the playbook

Troubleshooting

IssueResolution
Scanned PDFs with poor OCR qualityRe-process with enhanced OCR; flag low-confidence extractions for manual review
Contract is an amendment with no standalone termsLink to parent agreement; extract only the modified terms and note inherited provisions
Ambiguous or missing provisionsMark as "Not Found" or "Ambiguous" — never guess; flag for human review
Conflicting terms between parent and amendmentThe amendment controls unless it explicitly preserves the original; flag the conflict
Very large corpus (10,000+)Process in batches by contract type or counterparty; produce interim reports
Playbook doesn't cover a provision found in contractsNote the gap; suggest adding the provision to the playbook

Individual skills in this repo

This repo contains 20 individual skills — each has its own dedicated page.

ThomasMoreAI/legal-skills-open

Track the IP portfolio — registrations, renewals, maintenance fees, and use declarations. Use when checking what's renewing, adding or updating an asset, recording a maintenance filing, or auditing the register for gaps, lapses, and use-in-commerce questions. Receives handoffs from prosecution and clearance work.

ThomasMoreAI/legal-skills-open

When the user's message starts with "Canada" (case-insensitive), generate a full Ontario PI Case Assessment Memo from the case_stressor corpus and return it verbatim. This is the primary handler for Canadian fact patterns — DO NOT ask clarifying questions first; run the memo and let the lawyer react.

ThomasMoreAI/legal-skills-open

交通事故责任评估与判定专业技能。当用户需要交通事故责任分析、事故现场照片评估、交通法规咨询、事故责任划分、法律依据查询时触发此技能。适用于车辆碰撞事故、行人事故、非机动车事故等各类道路交通事故的责任认定场景。无论用户使用"交通事故"、"车祸"、"责任判定"、"交通法规"、"事故定责"等何种表述,只要涉及交通事故评估或责任认定,均应调用此技能。

ThomasMoreAI/legal-skills-open

Strategische Markenportfolio-Planung für Luxus-Modehaeuser: Mandant will Marken in DE/EU/international schützen oder Portfolio optimieren. Normen: §§ 32 ff. MarkenG, Art. 32 ff. UMV (EU) 2017/1001, Madrid-Protokoll (WIPO). Prüfraster: Nizza-Klassen (3/14/18/25/35), Multi-Class-Strategie, Prioritaets-Kaskade, Kostenoptimierung, Anmeldezeitpunkt. Output Marken-Portfolio-Plan, Anmelde-Empfehlung je Territorium, Kostenprojektion. Abgrenzung: Einzelne Anmeldung DPMA siehe wortmarke-anmeldung-dpma; Madrid-Protokoll Details siehe madrid-protokoll-und-internationale-registrierung.

ThomasMoreAI/legal-skills-open

Audit an intellectual-property portfolio for ownership, protection, scope, deadlines, territorial coverage, use, value, encumbrances, and enforcement risk. Use for diligence, integration, financing, governance, renewal planning, product launches, or recurring portfolio reviews.

ThomasMoreAI/legal-skills-open

Use whenever the user asks about a motor-vehicle statute, citation, contributing factor, OR a Canadian personal-injury fact pattern — always query Specter's Harvester API before answering. The API auto-routes between two collections (US statutes + Canadian PI case law).

ThomasMoreAI/legal-skills-open

Assesses product liability exposure on given facts — classifying the defect as manufacturing, design, or warning/instruction, mapping which party in the supply chain is potentially exposed, and grading the realistic exposure. Use this whenever a user wants product liability worked through rather than a general deficiency test — including phrasings like "what's our exposure if this product injured someone", "is this a design defect or a manufacturing defect", "who in the supply chain is on the hook here", "assess our product liability risk on these facts", or "how exposed are we if the warning label was inadequate". Distinct from deficiency-analyst, which tests service and trade-practice thresholds — this is specific to defective products and supply-chain exposure. Fires for any product liability question, in any jurisdiction, for manufacturers, assemblers, sellers, distributors, or importers.

ThomasMoreAI/legal-skills-open

KI-VO Hochrisiko-Anforderungen für Personalwesen in Kanzleien ab August 2026: Anwendungsfall Kanzlei setzt KI im HR-Bereich ein oder beraet Mandanten zum AGG-konformen KI-Einsatz bei Bewerberauswahl. Anhang III Nr. 4 KI-VO Hochrisiko Bewerberauswahl, Inkrafttreten 2. August 2026, AGG Diskriminierungsverbot. Prüfraster Hochrisiko-Klassifizierung eigener HR-KI, Konformitätsbewertung, Transparenzpflichten für Betroffene, Beratungsmandate Arbeitsrecht. Output Checkliste Hochrisiko-Anforderungen mit Umsetzungsplan für August 2026. Abgrenzung zu Bias-und-Diskriminierung-Prüfung und zu KI-VO-Betreiber-Pflichten.

ThomasMoreAI/legal-skills-open

Vorlagetabelle für Portfolio-Review von Arbeitsvertraegen im 3D-Format: Forderung/Prüfung/Stellung. Normen: BGB, KSchG, ArbZG. Prüfraster: Vertragsbedingungen, Klauselgueltigkeit, HR-Compliance. Output: Arbeitsvertrag-Portfolio-Tabelle. Abgrenzung: nicht allgemeine 3D-Review-Konfiguration.

ThomasMoreAI/legal-skills-open

Unternehmen oder Kanzlei muss IP-Portfolio verwalten und anstehende Fristen im Blick behalten. Schutzrechtsportfolio-Verwaltung. Prüfraster: Eintragungen Verlaengerungen Jahresgebühren Benutzungsnachweise Fristkalender. Output: Fristenkalender und Portfolio-Audit mit Luecken Verfall und Benutzungsfragen. Abgrenzung zu schutzschrift-eilverfuegung (Verletzungsverteidigung) und markenanmeldung-dpma.

ThomasMoreAI/legal-skills-open

Statusuebersicht aller laufenden Prozessmandate: Fristen, Verfahrensstand, naechste Schritte. Normen: ZPO, RVG. Prüfraster: Fristenliste, offene Anträge, Termine, Mahnfristen. Output: Portfolio-Statusbericht Prozessmandate. Abgrenzung: nicht Einzelmandat-Briefing.

ThomasMoreAI/legal-skills-open

Workflow-Skill zu aufklaerungsfehler beweisstrategie. Nutzt Normtext, Nutzerangaben und verifizierte Quellen; Rechtsprechung nur nach Live-Pruefung mit Gericht, Datum und Aktenzeichen.

ThomasMoreAI/legal-skills-open

Strukturierte Prüfung von Ansprüchen wegen Behandlungsfehler nach §§ 630a ff. BGB iVm § 823 BGB. Behandlungsvertrag Aufklärungspflicht § 630e BGB Dokumentationspflicht § 630f BGB Beweislastregeln § 630h BGB grober Behandlungsfehler Beweislastumkehr voll beherrschbares Risiko Anfaengerstandard Schmerzensgeld § 253 BGB. Schlichtungsstelle Aerztekammer MDK-Gutachten. Verjährung drei Jahre § 195 BGB Hoechstfrist dreissig Jahre § 199 Abs. 2 BGB.

ThomasMoreAI/legal-skills-open

Workflow-Skill zu fachanwalt medizinrecht aufklaerungsfehler. Nutzt Normtext, Nutzerangaben und verifizierte Quellen; Rechtsprechung nur nach Live-Pruefung mit Gericht, Datum und Aktenzeichen.

ThomasMoreAI/legal-skills-open

Behandlungsfehler §§ 630a 630h BGB Verletzung medizinischer Standard. Diagnosefehler Therapiefehler Befunderhebungsfehler Hygienefehler. Beweisregeln § 630h BGB Vermutung Kausalität bei grobem Behandlungsfehler § 630h Abs. 5 BGB Befunderhebungsfehler Dokumentationsmangel. Schadensersatzanspruch §§ 280 823 BGB Schmerzensgeld § 253 BGB. Verjährung drei Jahre § 195 BGB ab Kenntnis 30 Jahre Hoechstfrist.

ThomasMoreAI/legal-skills-open

Mandant hatte Verkehrsunfall und fragt: Wer haftet wie viel und welche Schadensposten koennen geltend gemacht werden? §§ 7 17 18 StVG iVm § 254 BGB Haftungsquote. Prüfraster: Betriebsgefahr beidseitig Anscheinsbeweis Auffahrunfall Spurwechsel Rotlicht Vorfahrt Mithaftung Tempo Sicherheitsabstand Anschnall. Schadenspositionen Reparatur fiktive Abrechnung Mietwagen Nutzungsausfall Sachverständige Schmerzensgeld. Output: Haftungsquoten-Berechnung und Schadenstabelle. Abgrenzung zu fachanwalt-verkehrsrecht-regulierungsanforderung (Gläubigerseite vs. Versicherer) und fachanwalt-verkehrsrecht-versicherer-quotenverhandlung-vergleich.

ThomasMoreAI/legal-skills-open

Verkehrsunfall-Mandat im Zivilprozess vorbereiten: Schadensersatz, Schmerzensgeld, Versicherungskorrespondenz. Normen: §§ 7 18 StVG, §§ 823 253 BGB, § 115 VVG. Prüfraster: Haftungsquote, Schadensposten, Verjaebrung, Regulierungsablauf. Output: Klageschrift Verkehrsunfall oder Klageerwiderung. Abgrenzung: nicht Strafrecht oder Ordnungswidrigkeiten.

ThomasMoreAI/legal-skills-open

Classifies personal vs non-personal data per GDPR Art. 4(1) definition test with decision tree for borderline cases. References Breyer v Germany CJEU C-582/14 dynamic IP ruling and WP29 Opinion 4/2007. Keywords: personal data, GDPR Art 4, data classification, Breyer ruling, identifiability test, PII.

ThomasMoreAI/legal-skills-open

Roll up the portfolio from _log.yaml — risk distribution, upcoming deadlines, stale matters, materiality totals, stage distribution, and flagged anomalies.

ThomasMoreAI/legal-skills-open

Use when a matter involves marriage, divorce, property division, custody, or inheritance for non-Muslim residents of the UAE under Federal Decree-Law 41/2022 (the UAE Civil Personal Status Law) or the DIFC Wills Service Centre. Covers no-fault divorce, joint-custody default, equal-property division, testamentary freedom for non-Muslims, pre-nuptial agreement enforceability, and the parallel DIFC/ADGM wills regimes. Contrasts with the UAE Personal Status Law 28/2005 (Sharia-based) which governs Muslim residents. Triggers on non-Muslim divorce UAE, UAE civil marriage, DIFC wills, UAE prenup, or expat family law UAE questions.

相关技能