> For the complete documentation index, see [llms.txt](https://help.craftsmanplus.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.craftsmanplus.com/intelligence/chatbot/user-input.md). # User Input How to write to CRAFTSMAN INTELLIGENCE: the message box, sending, steering a response in progress, stopping, editing, and starting a new chat. Everything you say to CRAFTSMAN INTELLIGENCE starts in the message box at the bottom of the chat. This page covers typing and sending messages and the controls around them. The pages in this section cover the other ways you can give the assistant input: * [**Attachments**](/intelligence/chatbot/user-input/attachments.md) — Add files and Asset Library assets to a message, and reuse media already in the conversation. * [**Questions**](/intelligence/chatbot/user-input/questions.md) — Answer when the assistant asks you to choose between options. * [**Quick Replies**](/intelligence/chatbot/user-input/quick-replies.md) — Send a follow-up or a template prompt with one click. *** ### The Message Box Type your request in the message box. There is no special syntax to learn. Write the way you would describe the task to a colleague, and the assistant works out which specialty should handle it. On a new chat, the empty message box cycles through example prompts from the specialties available on your current page. This is only a preview of what you can ask. It stops as soon as you start typing and does not return once the chat has its first message. * Press **Enter** to send. * Press **Shift + Enter** to add a line break. * The box grows as you type. Long drafts scroll once it reaches its maximum height. There is no length limit on a message. *** ### Sending a Message Click **Send** or press **Enter**. The **Send** button is available as soon as your message has something in it: text, an attachment, an Asset Library asset, or a selected image or audio clip from earlier in the conversation. A message does not need text if it has an attachment. {% hint style="info" %} *If you have just added a file, **Send** stays disabled until the upload finishes. Wait for the spinner on the file tile to clear, then send.* {% endhint %} While the assistant works, the chat shows a thinking indicator that updates with what it is doing, such as *"Generating 3 images"* or *"Searching the web"*. Click the indicator to expand the steps. *** ### Enhancing Your Prompt Click the **wand** button next to the message box to rewrite your draft into a more detailed version. The assistant expands your request with the specifics a designer would assume, without inventing facts you did not provide. The rewrite replaces the text in the message box. Review it, edit it if you like, and send when you are ready. The button is disabled while the box is empty, and typing while a rewrite is in progress cancels it. *** ### Adding to a Response in Progress You do not have to wait for the assistant to finish before sending your next thought. While a response is in progress the message box stays open with the placeholder **Add to steer the agent...**, and the **Send** button tooltip reads **Queue message to steer the agent**. Messages you send during a response are **queued**, not sent immediately. They appear below the conversation at reduced opacity so you can tell the assistant has not seen them yet. The assistant picks them up between its steps and takes them into account before continuing. You can attach files to a queued message, and you can remove a queued message with the **X** that appears when you hover over it. {% hint style="info" %} *Queued messages are a way to redirect the assistant mid-task, for example "use blue instead of red" while it is still building. They do not interrupt what it is doing right now.* {% endhint %} *** ### Stopping a Response Click **Stop** to end the current response. The **Stop** button appears next to **Send** whenever the assistant is working. When you stop: * Whatever the assistant has already written stays in the chat, followed by the line **You stopped this response.** * Any queued messages are discarded. * Changes the assistant already made to your project are kept. Stopping is not an undo. To reverse a change, use the editor's **Undo**, which reverts the whole response's changes in one step. *** ### Editing a Sent Message Hover over one of your own messages and click the **Edit** (pencil) icon to change it. Edit the text, then click **Save** or press **Enter**. Press **Esc** or click **Cancel** to leave it unchanged. Saving an edit removes every message after the one you edited and sends the edited message again, so the assistant responds fresh. Any files attached to the message are kept. {% hint style="warning" %} *Editing rewinds the conversation. Messages and responses after the edited one are removed and cannot be recovered.* {% endhint %} You can also hover over any message and click the **Copy** icon to copy its text. *** ### Starting a New Chat Click the **New Chat** (+) button in the chatbot header, or press **⌘ Shift O** on Mac or **Ctrl + Shift + O** on Windows. The shortcut works from anywhere in the platform, even when the chatbot is closed, and opens it on a fresh chat. Your previous chat is saved in [History](/intelligence/chatbot/history.md). A new chat also starts automatically when you open a different project. *** ### Rating Responses Hover over any response from the assistant to show three actions: **Good Response**, **Bad Response**, and **Copy Response**. Ratings help us improve the assistant. Click the same rating again to remove it. *** ### If Something Goes Wrong If a message fails, the chat shows the reason in a red notice with a **Dismiss** button and, when it makes sense to try again, a **Retry** button that sends your last message again. Connection problems, rate limits, and temporary server issues can all be retried. If the AI provider itself is unavailable, the notice reads **Intelligence Unavailable** and chat resumes once service is restored. If the chatbot is closed or your tab is in the background when a response finishes, a sound plays and the **CRAFTSMAN INTELLIGENCE** button and the browser tab icon show that there is something new to read. --- # 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.craftsmanplus.com/intelligence/chatbot/user-input.md?ask=&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.