Guide 26About 8 minUpdated 2026-06-22New

Hide Settings and Open Them by Volume: You Control What AI Sees

Don't want AI spoiling the culprit or the final twist? Hide settings, characters and suspense from AI, or open them volume by volume. In Volume 1 the AI sees only the surface identity; the truth behind it unlocks in Volume 3. Single-item and batch demos included, so every reveal lands on cue.

For: Authors with twists and foreshadowingFor: Long-form authors worried that AI will spoil the plot

The product interface is currently in Chinese, and the screenshots in this tutorial are shown as they are. Please follow the screenshots for button and menu names.

Hide Settings and Open Them by Volume: You Control What AI Sees
In this article 6 sections

When you reach a key twist, you certainly don't want AI to blurt out the identity of the mastermind ahead of time; some foreshadowing has to wait until Volume 3, and the earlier chapters must never let AI know about it. Setting visibility is made for exactly this: you can hide a single setting, character or suspense item from AI completely, or let it open up to AI gradually, "by volume".

These switches are tucked into the reference sidebar on the left of the writing page. Click Settings at the top and you will see the settings, characters and suspense of the current book.

The settings sidebar

Setting visibility for a single entry

Open any setting (characters and suspense work the same way), scroll down in the detail card and find the ❖ AI visibility block.

The AI visibility switches of a single setting

There are two switches here:

  1. Hide from AI: once on, this entry is not assembled into any AI context. Even if "Settings" is turned on in context assembly, the hidden entry is skipped. It is the place for things that must never be spoiled: the true culprit's identity, the final ending, a hidden cheat ability.
  2. Open progressively by volume: once on, this entry is normally invisible to AI, and is unlocked for AI automatically only when your writing progress reaches the volume you specify.

How to use progressive opening by volume

After you turn on Open progressively by volume, all the volumes of this book are listed below so you can tick Start volume.

Multi-select of volumes for progressive opening

To understand the logic, remember one sentence: what you tick is "from which volume it becomes public". When your writing progress reaches the earliest volume you ticked, this setting automatically becomes visible to AI.

An example: you tick Volume 2 for "the heroine's true origins". While you write Volume 1, AI knows nothing about it and cannot leak it early; once you reach Volume 2, this origin setting automatically enters the context and AI starts writing according to it. No spoilers, and no need to flip the switch by hand every volume.

The 2 selected at the top right counts how many volumes you ticked. If you tick none, the entry stays hidden — the panel also reminds you of this in orange text below.

What the little badge on the card means

Cards with visibility settings show a small badge in the top right corner:

  1. A crossed-out eye icon: this entry is Hide from AI
  2. An open book icon: this entry is Open progressively by volume

A glance at the sidebar tells you what is hidden from AI, without opening entries one by one.

Changing many at once: batch visibility

To handle many entries in one go, click the multi-select icon at the top right of the sidebar to enter multi-select mode, tick the cards to process, and an AI visibility batch toolbar appears above.

Batch visibility in multi-select mode

There are three buttons:

  1. Hide: hide all selected entries from AI
  2. Open: make all selected entries visible again
  3. Open by volume: after clicking, choose the start volume and set progressive opening for all selected entries at once

When you tidy up the foreshadowing ledger of a volume, this is far faster than clicking one by one.

How it relates to context assembly

The two work together; just remember this priority: visibility is the "red line", context assembly is the "quota".

  • Context assembly decides "whether to give this category of material and how much" (for example, full settings or names only).
  • Setting visibility decides "whether this specific entry may be given".

Hidden entries, or entries whose opening volume has not been reached, never enter the context no matter how context assembly is set. So feel free to turn on "full settings" in assembly, and leave the real spoiler points to visibility as the safety net.

Once you control "what AI gets to see", continue with:

  1. Show the AI Only What It Needs: Context Assembly Settings
  2. Keep Characters From Forgetting and Settings From Contradicting
  3. Managing Characters, Worldbuilding and Suspense: The Left Sidebar
#258 min readUpdated 2026-06-22
New

Show the AI Only What It Needs: Context Assembly Settings

How much does the AI read, and how many credits does it burn? Configure context assembly item by item: full or names-only settings, outlines on demand, live token estimate—so the same credits stretch tens of thousands of words further, with less drift.

Authors who care about credit costsAuthors of long serials who need to control context
Read the tutorial
#147 min readUpdated 2026-03-24

Keep Characters From Forgetting and Settings From Contradicting

Use the knowledge graph to catch forgotten settings and continuity errors and check consistency, so a 300,000-character novel stays on track.

Long-form authorsAuthors who update frequently in serial publication
Read the tutorial
#127 min readUpdated 2026-03-24

Managing Characters, Worldbuilding and Suspense: The Left Sidebar

The left sidebar of the writing page manages characters, settings and suspense in one place: how to find it, create cards and look things up with floating cards.

Authors building a world and character filesAuthors who have started the main text and whose materials keep getting scattered
Read the tutorial
#279 min readUpdated 2026-06-22
New

Custom Agents: Give Each AI a Persona and Its Own Knowledge Base

Custom agents: four persona fields, 'gold-medal editor' prompt, agent/mode binding, one-click official templates to adapt—same prompt, new style.

Authors who want a fixed AI writing styleAuthors who need de-AI tone or a specific style
Read the tutorial