> 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/chat-agents-intangible/limits.md).

# Tips for directing agents

Habits that get better results from Intangible agents: specific prompts, plain character names, well-named models, undo, checking a video's cost, and asking a stalled agent to continue.

These habits get you what you meant the first time.

* **Be specific.** A short, vague request leaves the agents guessing, and sometimes they guess wrong. Name the subject, the place and the shot you want. If a scene change comes back off, undo it with Cmd/Ctrl+Z and ask again with more detail.
* **Give characters plain names.** For a character that must stay the same across shots, give it a plain name and put [image references](https://help.intangible.ai/learn-more/concepts/image-reference.md) on it, rather than describing it each time.
* **Name your models.** The agents find imported models by their names and descriptions. A model called "object-12" won't answer to "the green Bronco", so name and describe models when you import them.
* **Know what undo covers.** Cmd/Ctrl+Z undoes scene changes the agents made, in any mode. It doesn't undo canvas card actions, such as placing uploads or adding cards to a group, and the undo history clears when you reload the page. Remove an unwanted card by hand.
* **Check a video's cost before you agree.** The agents pick the video model themselves. Before they start a video, they say what you'll get and its credits, sometimes with one alternative, and wait for you to agree. A video the agents start doesn't show the cost notice the Visualize panel shows above **Generate**, so read their figure, or ask for the cost first. To use a particular model, name it.
* **If an agent seems to have stopped, ask.** Type "continue", or ask where it got to, for example "What is left on the step list? Continue from there." During a long build the chat panel also shows a **continue if you timed out** link.

For canvas limits that are not about the agents, see [Known limitations](https://help.intangible.ai/learn-more/faq/known-limitations.md). If a run fails, see [Troubleshooting Intangible agents](https://help.intangible.ai/learn-more/faq/troubleshooting-the-agent.md).

## Related

* [What the agents do in Build, Compose and Visualize](https://help.intangible.ai/chat-agents-intangible/in-the-modes.md)
* [Work well with Intangible agents](https://help.intangible.ai/chat-agents-intangible/work-well-with-intangible-agents.md)
* [Status and steps](https://help.intangible.ai/chat-agents-intangible/status-and-steps.md)


---

# 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/chat-agents-intangible/limits.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.
