> ## Documentation Index > Fetch the complete documentation index at: https://vaquill.ai/docs/llms.txt > Use this file to discover all available pages before exploring further. # Building a Playbook > Create a playbook, add clause rules, test-drive it on a sample draft, and compare it against market standard A playbook is a set of clause-level rules. The drafter consults it when generating contracts, and review uses it to score counterparty paper. This page covers creating one and working in the editor. For what each individual field does, see [Clause Rules](/docs/guides/playbooks/clause-rules). ## Create From the Playbooks page, choose **New playbook**. The form opens with the template gallery and any recent imports underneath it, so you can still switch to a faster start. **Name** is required. Make it specific enough that a teammate picking from a list knows what it is for, for example "SaaS vendor MSA, house playbook". **Contract type** determines which drafts and reviews this playbook can drive, and which playbook is resolved as the default for that type. **Description** is optional and is the right place for notes on when to use this playbook, its scope, and known exceptions. Pick clause types from the standard list, or define your own. Each one becomes a card you can open and fill in. Choose **Create playbook**. From here on, every save writes a version snapshot, so nothing you do in the editor is unrecoverable. From the playbooks list, mark it as the default for its contract type. Drafting and review then use it automatically when no playbook is picked explicitly. ## Adding clause rules The editor has two ways to add a rule. * **Standard clause types.** Pick from the built-in list. These are the clause types the rest of the product understands, so a standard type lines up with clause-library entries, default risk weights, review scoring, and the health analytics. * **Custom clause types.** Give the clause a name of your own, for example "Most favoured nation" or "Service levels". Use this only when nothing standard fits, because a custom type will not carry a default risk weight or match against the market-standard baseline. Each rule can be disabled without being deleted. A disabled rule is ignored by drafting, review, the structure check, and health analytics, but its text is kept so you can turn it back on later. ## Simple and Advanced The detail-level toggle at the top of the clause list controls how much of each rule is shown. | Mode | Shows | | ------------ | --------------------------------------------------------- | | **Simple** | Preferred position and acceptable range only | | **Advanced** | The full three-tab rule: Language, Guardrails, Governance | Simple is the right place to start. Write the wording you want for the ten or fifteen clauses you actually negotiate, save, and use it. Add depth later, clause by clause, as you learn where the arguments happen. The editor opens in Advanced automatically if the playbook already carries fallback ladders, walk-away floors, approvals, conditional rules, or numeric limits. That way a cloned template never hides its own depth from you. ## Test-drive it before you rely on it **Run sandbox** generates a small fixture draft using this playbook so you can see whether your positions actually produce the language you expect. It picks a handful of the highest-stakes clauses your playbook covers, drafts those sections against a sample deal, and scores each one back against the same playbook. You get, per section: * The drafted clause text. * A verdict of **Preferred**, **Fallback**, or **Needs attention**. * An overall risk band: **No covered sections**, **Low risk**, **Needs review**, or **High risk**. You can replace the sample fact pattern with your own before running. The default is a generic two-party agreement with an effective date. A sandbox run that comes back **Needs attention** on a clause you thought was well specified usually means the preferred position is written as an aspiration rather than as language. "Reasonable liability cap" is not draftable. "Mutual cap at 12 months of fees paid" is. ## Compare against market standard The **vs. market standard** panel puts your positions side by side with Vaquill's built-in baseline for the same clause and your active jurisdiction. It answers a question a rule sheet cannot: where has this team gone harder, or softer, than typical market terms? The comparison covers clauses in both, clauses only you have, and clauses only the baseline has. It is hidden for M\&A and for the specialized addenda, because there is no meaningful generic baseline for those. ## The three tabs | Tab | What lives there | | ------------ | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | **Editor** | Identity fields, the clause rule list, and the market-standard comparison | | **Insights** | The [structure check](/docs/guides/playbooks/structure-check), [playbook health](/docs/guides/playbooks/playbook-health), and [drift suggestions](/docs/guides/playbooks/drift-suggestions), all reading the saved playbook | | **History** | [Version snapshots](/docs/guides/playbooks/versions-and-sharing), with diff and restore. Appears once there is at least one snapshot | ## Limitations * Insights read the saved playbook, not your unsaved edits. Save before checking whether a change fixed a finding. * The market-standard comparison is not available for every contract type. * A sandbox run costs a model call and drafts only a few sections. It is a smoke test, not a review of the whole playbook. * Custom clause types work, but they sit outside the standard vocabulary, so they get no default risk weight and no market-standard comparison. * Only the owner can edit. Teammates with access can read and use the playbook, and must duplicate it to make changes. * There is no approval flow on the playbook itself. Anyone who owns it can change a position and it takes effect on the next draft or review. ## Related Every field on a clause rule and what it changes downstream. Clone a curated playbook instead of starting from a blank page. Catch authoring gaps before the playbook sees a deal. Snapshots, restore, defaults, and team access.