> For the complete documentation index, see [llms.txt](https://help.intangible.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.intangible.ai/canvas/what-intangible-agents-build.md).

# What Intangible agents build

What lands on the canvas after you confirm the brief: reference sheets, generated references, a scene at true scale with its models and motion, and a shot card for each shot, each one a card you can o

You bring a prompt and maybe a few photos. A shot needs much more: references for every subject, 3D models, a set at the right scale, motion, and a camera. Once you confirm the brief, the agents fill in what is missing and show each piece as it lands.

In one run of the airbag rover example, two crater photos and one paragraph became three reference sheets (pod, rover, crater), a lunar crater scene with the pod, lander and rover inside it, the drop and bounces animated at lunar gravity, and five framed shots.

## What they make

**Reference sheets.** For each subject group the agents made from your images, they start a reference sheet: one clean image that keeps the subject looking the same in every shot. The sheet lands in its group, marked as a reference sheet, and replaces the loose photos as the subject's reference. They skip any group that already has one, including one you uploaded. All of a project's sheets are drawn in one style, taken from the brief's look, so the characters, objects and locations read as one piece; ask for a different style on a sheet and they use yours. Change the project's style later and the sheets already made keep the old one, so ask the agents to remake them. Each sheet costs credits.

**References for what you did not bring.** If the brief needs something you gave no picture of, the agents generate a reference image for it. Each shows in the step list as it is made.

![Groups filling in as the agents work: the Characters and Cars placeholders turn into reference sheets, then Las Vegas Strip at night, Casino and Vegas Hotels fill with generated references (sped up)](https://1179478100-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FSWrjIqlgVAblx6vHGJiM%2Fuploads%2Fgit-blob-fbbf8f7ea5546aa4654a391cd8ea6a5231983760%2Fcanvas-groups-fill-01.gif?alt=media)

**The scene and its models.** They build the set at true scale and place a model for each subject in it. The models live inside the scene: they do not get cards of their own on the canvas. The scene card in **Storyboard** shows a live preview of the set.

**Motion.** Where the brief calls for it, the agents animate the scene, for example a pod dropping and bouncing under lunar gravity.

**Cameras and shots.** They frame a camera for each shot in the brief, and a shot card for each appears in **Storyboard** next to the scene card, for example 01 Cut loose to 05 Drive-off. A shot card always shows a live view through its camera at the shot's first frame, so you can compare it with the renders made from it.

![The Storyboard group filling scene by scene with shot cards, then the 3D set behind those shots (sped up)](https://1179478100-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2FSWrjIqlgVAblx6vHGJiM%2Fuploads%2Fgit-blob-6ae5a71ab148e287d74191f862edcac189fa20e1%2Fcanvas-storyboard-to-build-01.gif?alt=media)

## When they finish

With Guide me, the agents stop before rendering anything. Their report lists what is in the scene, the motion, each shot with its status, and any problems they could not solve, such as a sun direction they could not check. Then a choice card asks **What next?** with options such as fixing some shots first or rendering all of them. Answer it, or ask for a fix first, before anything renders. Check the canvas for the cards before you move on.

## Change any of it

Everything the agents build is ordinary project content:

* **Ask for a change** in the chat panel, and name the card you mean. See [Chat with Intangible](https://help.intangible.ai/chat-agents-intangible/chat-with-intangible.md).
* **Open it in a mode.** Select a scene card and click **Edit in Build** to change the set; select a shot card and click **Edit in Compose** to change the framing. See [Take control](https://help.intangible.ai/get-started/take-control.md).

## Put your own models on the canvas

A model you add in Build gets a card on the canvas once it has a reference image attached. To show any other model, select it in Build, open **More options** and turn on **Promote to Canvas**. Select a whole group of models to show it as one group card. Turn the option off to take the card away.

## Related

* [The brief](https://help.intangible.ai/canvas/the-brief.md)
* [Status and steps](https://help.intangible.ai/chat-agents-intangible/status-and-steps.md)
* [Groups and cards](https://help.intangible.ai/canvas/groups-and-cards.md)
* [Batch visualize](https://help.intangible.ai/canvas/batch-visualize.md) – the next step: a keyframe for every shot.


---

# 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://help.intangible.ai/canvas/what-intangible-agents-build.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.
