Clause checklist - what a contract is missing, and what bites

Philosophy

The most expensive clauses are the ones that are not in the contract. Reviewing a contract is not only reading what is written - it is checking that no safeguard this deal needs is absent. This skill walks the contract against a fixed list of 41 clause categories and, for each, says: present / absent / present-but-risky, quoting the clause when present.

It is extractive - what it shows comes from the contract, not from paraphrase. It flags presence and risk; the assessment and the call stay with the lawyer.

One caution built into the method: CUAD is an extraction taxonomy, not a list of safeguards every agreement must contain. Several categories are deal-specific, and some are mutually exclusive pairs (uncapped vs cap on liability; unlimited vs limited licence grants). Treating every absence as a red flag buries the real gaps in false positives.

The 41 categories (CUAD taxonomy)

Groups follow CUAD (1-6) plus ungrouped categories. Two steps per category: first applicability - does this deal need the clause, given the contract type, the parties and the subject matter; then, for applicable categories only: present / absent / risky + a quote when present. Mutually exclusive pairs count as one decision - record which side the contract takes, not the "absence" of the other side.

Contract metadata

• Document name, Parties (verify authority to sign), Governing law.

Group 1 - term and dates

• Agreement date · Effective date · Expiration date · Renewal term (auto-renew?) · Notice period to terminate renewal.

Group 2 - competition restrictions

• Non-compete · Exclusivity · No-solicit of customers · Competitive restriction exception (carve-outs).

Group 3 - control and assignment

• Change of control (consent/termination on M&A?) · Anti-assignment.

Group 4 - licences

• License grant · Non-transferable license · Affiliate license (licensor) · Affiliate license (licensee) · Unlimited/all-you-can-eat license · Irrevocable or perpetual license.

Group 5 - post-term and audit

• Post-termination services · Audit rights.

Group 6 - liability

• Uncapped liability · Cap on liability.

Ungrouped

• Most favored nation · No-solicit of employees · Non-disparagement · Termination for convenience · ROFR/ROFO/ROFN · Revenue/profit sharing · Price restrictions · Minimum commitment · Volume restriction · IP ownership assignment · Joint IP ownership · Source code escrow · Covenant not to sue · Liquidated damages · Warranty duration · Insurance · Third-party beneficiary.

Output format

RED FLAGS (top):
- ABSENT: <category, applicable to this deal> - <why this deal needs it>
- RISKY: <category> - <how it is one-sided>
(a category marked not applicable never becomes a red flag)

41-CATEGORY TABLE:
| Category | Applicable | Status | Note | Quote (if present) |
| Governing law | yes | present | New York; courts of NY | "..." |
| Cap on liability | yes | absent | no liability ceiling | - |
| Source code escrow | no - no software licensed | n/a | - | - |
...

Limits

• A presence-and-risk checklist, not an interpretation of clause wording. Whether a clause is effective, and the recommendation, are the lawyer's.
• The taxonomy is common-law-oriented at source (CUAD); for a specific jurisdiction add local-law anchors (for Poland, see klauzule-kontraktowe-pl).
• Descriptive references without a clause ("the parties intend to cooperate") have nothing to match - mark as "no clause, manual check".

Attribution

The 41-category clause taxonomy is based on CUAD (Contract Understanding Atticus Dataset), The Atticus Project, CC BY 4.0 (https://www.atticusprojectai.org/cuad). Category descriptions and the risk framing are MateMatic's own. MateMatic interpretation, not the position of any bar or regulator.

Shared rules of the verification-foundation plugin

These rules apply to every skill in this plugin, including where the skill itself is silent. They are copied into each skill, so they hold whether you install the whole plugin or a single skill.

Five layers of protection (before any disclaimer)

1. Source verification - provisions and case law from databases, not the model's memory.
2. Confidence class - every legal claim marked: verified / check / do not use.
3. Premise check - facts from the user are tested before analysis.
4. Explicit negative scope - each skill states what it does NOT do.
5. Human gate - a qualified person reviews and approves the output.

If an error would pass without being stopped by layers 1-5, the fault is in the skill. Fix the tool, do not add a note. "This is not legal advice" stops no error and transfers no responsibility.

Confidence tags

• Verified - source checked in this session, with a full identifier: (GDPR Art. 6), (Case C-311/18, CJEU).
• Check - plausible, unverified: [check on EUR-Lex].
• Do not use - an invented case number or provision. Omit it, never fabricate.

The tag sits next to the line it concerns. A case number that exists is not enough - check the content.

Human gate

Nothing is sent, filed, signed or published before a qualified person checks and approves it. A skill output is a draft, not a finished document.

Scope

The verification foundation is jurisdiction-neutral - the method, not the substance of any one legal system. It connects to no external source (no MCP connectors) and makes no outbound calls of its own; what you type still goes to your configured model (see https://github.com/matematicsolutions/awesome-matematic-skills-en/blob/main/TRUST.md). Substance comes from separate plugins (eu-law-sources). Deep national law (e.g. Polish case law) lives in the Polish marketplace, by design.