Every legal team has positions it takes: what it always asks for, what it will trade away, and what it will never accept. A playbook records those positions as rules. When an AI review runs against it, the contract is compared to your standards rather than to a generic idea of a good contract — so two people reviewing two contracts on two days reach the same conclusion.
A worked example
Suppose your payment rule says:
- Standard
- Invoices should be paid within 30 days.
- Fallback
- Payment within 45 days is acceptable for large customers.
- Unacceptable
- Payment after 60 days or more needs Finance approval.
Now someone uploads a contract with a 90-day payment term. The review flags that clause as unacceptable automatically, names the rule it broke, and the reviewer sees the risk without having to remember what your policy was.
How a playbook is built
- Playbook
- One rule set for one kind of agreement. It carries a contract type, a jurisdiction, a workspace scope and a status.
- Sections
- The areas of the contract. A new manual playbook starts with General provision, Payment, Termination and Governing Law, and you can rename these or add your own.
- Rules
- One negotiating position each, with a number, a title, a severity and up to three kinds of position text.
The three positions in a rule
Each rule describes what you want and how far you will move. This is what lets a review tell you not just that a clause is unusual, but whether you can live with it.
- Standard position
- The preferred negotiation outcome — the wording you would put in your own draft.
- Fallback positions
- Acceptable alternatives, ordered from best to worst. The first is your first concession, the next comes after that.
- Unacceptable positions
- Blockers and escalation points. When a contract matches one of these, the review says so plainly.
Only the standard position is needed to make a rule useful. Fallbacks and unacceptable positions make the findings sharper, but you can add them later as real negotiations reveal where your limits actually are.
Rule severity
Severity tells a reviewer how hard to push. Four levels are available:
- Blocker
- Do not sign while this is unresolved.
- High
- Material risk — escalate before agreeing.
- Medium
- Worse than standard; negotiate where you can.
- Low
- A preference rather than a requirement.
Scope, language and status
- Workspace scope
- A playbook is either company-wide or scoped to a single workspace. Scope it to a workspace when one team negotiates on terms the rest of the company does not use.
- Language
- Set from your interface language when the playbook is created, and shown in the playbook list. Rules are written in that language.
- Status
- Draft while you build it, Active once it is ready. Only active playbooks can be selected in a review.
Common questions
Do I need a separate playbook for each contract type?
Yes, and that is the point. An acceptable liability cap in an NDA is nothing like one in a supply agreement, so each contract type gets its own rule set and its own selection at review time.
What does the Rules indexed number on the playbook list mean?
It is the total number of rules across all your playbooks — a quick measure of how much of your negotiating policy is actually written down.
Can a playbook be in Turkish?
Yes. A playbook takes its language from your interface language at the moment it is created, so switch to Turkish first if you want Turkish rules. Many teams keep an English and a Turkish playbook for the same agreement type.
Who can see a playbook?
Anyone who can reach its scope. A company-wide playbook is visible across the company; a workspace-scoped one only to members of that workspace.
About this guide
Imported from the application Help Center on 23 September 2026. The original content date is preserved. Product workflows and screenshots have not been revalidated against current account permissions and availability.