> For the complete documentation index, see [llms.txt](https://docs.spicychat.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.spicychat.ai/product-guides/rulebooks/influence-rules.md).

# Influence rules

Influence rules make the character react to the state of the meters. The core mechanic: levels. Every instruction on this page goes straight into the prompt.

{% hint style="success" %}
**You don't need all of this.** Most Rulebooks use 1 or 2 Bar meters, 2 to 4 AI rules and 3 levels. Everything else on this page is optional. Start with a [Starter](https://docs.spicychat.ai/product-guides/rulebooks/examples-to-clone.md) and come back here when you want more.
{% endhint %}

## Levels

Levels split a Bar, Pips, Dice or Story clock meter into named zones. Each level can carry:

* Instruction: what the model should do while the meter is in this level ("are warm but not intimate", "is on edge and says little"). Bar, Pips, Dice, Story clock only. You can also write a rule that fires on a Pick one or Collection condition ([Rules](#pick-one-and-collection-rules) below).
* Status line: a short label users see as the caption under the meter ("Warming up", "On edge"). Without one, the level name shows.
* Color: only for Dice.
* Group: a short label for a family of levels ("Refusal", "Breakthrough"). The group name shows in the chat strip instead of the level name, so levels that read "Not", "Full" and "Over" can all show "Tank". Colors follow the group's color.

Example: a level at 50% of the range for Bar, Pips and Dice can be "fills a third", "fills two thirds" or "fills all". Pips steps map to levels in order. There is no level above the maximum: everything from the last bound to the max belongs to the last level.

Each level keeps a minimum segment width on the strip. With more than 8 levels the strip becomes a read-only overview and the levels are edited as a row list.

<figure><img src="https://3060264960-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FZHNBN3T5KDxoXWSU7L3d%2Fuploads%2F8uVoO099YXXEzmHtjUxO%2Frulebooks-editor-levels.png?alt=media" alt="Levels: zones, status lines, instruction fields" width="390"><figcaption><p>Levels: zones, status lines, instruction fields</p></figcaption></figure>

<figure><img src="https://3060264960-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FZHNBN3T5KDxoXWSU7L3d%2Fuploads%2FMHq2eR5SYLwtZl5kywN3%2Frulebooks-480-level-strip-crowded.png?alt=media" alt="More than 8 levels: a read-only strip overview with a row list below" width="390"><figcaption><p>More than 8 levels: a read-only strip overview with a row list below</p></figcaption></figure>

A Bar meter without levels sends no level instruction, and the meter shows no caption.

The Story clock has no zone strip. Its levels are set per dimension in Influence rules: Time of day (morning, afternoon, evening, night) and Story day (Day 1 to 2, "the first weekend"). At least one dimension stays enabled, and a Date only clock only offers Story day. Each dimension can carry its own instruction ("at night he gets bolder", "on Day 1 everyone is a stranger").

<figure><img src="https://3060264960-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FZHNBN3T5KDxoXWSU7L3d%2Fuploads%2FRLmG6NbX63ZtG1lr3MoG%2Frulebooks-480-clock-levels.png?alt=media" alt="Story clock dimensions: Time of day and Story day" width="390"><figcaption><p>Story clock dimensions: Time of day and Story day</p></figcaption></figure>

### Pick one and Collection rules

* Pick one: one instruction per item. While the item is selected, its instruction goes to the model and its caption shows as the status line (no Group option).
* Collection: rules with a condition (has at least, holds at least, doesn't have, holds nothing) and an instruction. While the condition matches, the instruction goes to the model.
* These are the only rules Pick one and Collection have: the conditions also work in [combined rules](#combined-rules), [triggers](https://docs.spicychat.ai/product-guides/rulebooks/making-meters-move.md#triggers), [Only when](https://docs.spicychat.ai/product-guides/rulebooks/making-meters-move.md#ai-rules) and dice.

<figure><img src="https://3060264960-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FZHNBN3T5KDxoXWSU7L3d%2Fuploads%2FA2xDrjXgqFOyjAmtLKlL%2Frulebooks-editor-pickone-rules.png?alt=media" alt="Pick one: a rule per item" width="390"><figcaption><p>Pick one: a rule per item</p></figcaption></figure>

### Journal rules

A [Journal](https://docs.spicychat.ai/product-guides/rulebooks/meters.md#journal) has rules with a condition and an instruction that goes to the model while the condition matches:

* recorded: the named event is in the journal.
* not recorded: the named event is not in the journal yet, for "they are still strangers" instructions.
* recorded at least: at least N records exist, of the named event when one is picked, in total when not.
* recorded nothing: the journal is empty, so the case or the story has not started.

An optional caption shows as the status line while the rule holds. [Journal timing](https://docs.spicychat.ai/product-guides/rulebooks/making-meters-move.md#journal-timing) conditions also work in combined rules, triggers and Only when.

## Combined rules

When a [level](#levels) cannot express what you need, combined rules send an instruction when several meter conditions are true at once. Up to 30 rules. Each rule has up to 6 conditions plus an instruction that applies while they hold.

Conditions can be:

* Meter vs a value: Affection above 50, Trust at most 20, Budget exactly 0 (at least, at most, above, below, equals).
* Story clock: time of day is, day is, day at least, day at most, weekday is, date is.
* Pick one: is or isn't one item.
* Collection: has at least N, holds at least 1, doesn't have, holds nothing.
* Journal: an event last recorded at least or at most N turns or story days ago.

Every condition on the same rule has to be true for the instruction to go to the model. Read it as one sentence: "While the place is Tavern AND suspicion is above 50, ..."

<figure><img src="https://3060264960-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FZHNBN3T5KDxoXWSU7L3d%2Fuploads%2F1qdMQSjIrsjAbmNGSHhx%2Frulebooks-editor-combined-rule.png?alt=media" alt="A combined rule: place is Tavern and Suspicion above 50" width="390"><figcaption><p>A combined rule: place is Tavern and Suspicion above 50</p></figcaption></figure>

## Visibility conditions

Visibility conditions control when a meter is seen in chat, not what the character does.

* A Hidden meter is never in the strip and never in the sheet: the user can't see it at all.
* Panel only meters stay out of the strip but keep a row in the [explanation sheet](https://docs.spicychat.ai/product-guides/rulebooks/using-a-rulebook-in-chat.md) and in move notes.
* Show when: hides a Shown or Panel only meter in chat while its conditions hold. It still works and it still shows for you in the editor and in [Playtest](https://docs.spicychat.ai/product-guides/rulebooks/playtest-and-checks.md).

Choose the meter, then the same conditions as combined rules.

***

**Up next:** Test your Rulebook before anyone else sees it.

{% content-ref url="/pages/gwyx5rXQRgo9G0zjuHB6" %}
[Playtest and checks](https://docs.spicychat.ai/product-guides/rulebooks/playtest-and-checks.md)
{% endcontent-ref %}


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.spicychat.ai/product-guides/rulebooks/influence-rules.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
