> For the complete documentation index, see [llms.txt](https://docs.lleverage.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.lleverage.ai/operator-agent/getting-started.md). # Getting Started The Operator Agent is accessible from the main navigation inside any project. Open the agent section to land on its home screen, where you can start a conversation or pick up a suggested task. ### The Agent Home Screen Before you start a session, the home screen offers a few ways in: * **Suggestion cards** - Cards suggesting tasks the agent can help with, tailored to your project. Pick one to start a session with that task, or ignore them and type your own message. * **Attention cards** - If the agent is waiting on you - for example a pending approval or an unanswered question from an earlier session - a card appears so you can respond without hunting for the right conversation. * **Onboarding cards** - When a project is new, checklist cards walk you through introducing yourself, connecting your tools, and giving the agent its first task. Builders can shape this screen for the whole team: 1. **Open the card editor** from the agent home screen. 2. **Drag cards to reorder** them, edit a card's content, or hide individual cards. 3. **Toggle whole sections** - hide the entire cards section or switch suggestions off altogether. ### Configuring Your Agent The agent's identity and behaviour are set per project: * **Name** - The project name is also the agent's name. By default the agent is called Llev. * **Project instructions** - A description of what the project is about, given to the agent at the start of every session. Use it for standing guidance: what the team does, preferences, things to always or never do. * **Agent tone** - How the agent sounds in replies: Crisp, Balanced, or Warm. * **Organisation branding** - Optionally show your organisation's logo on the agent's home screen. > 💡 Instructions and tone apply to every session in the project, so everyone on the team gets a consistently behaving agent. ### Your First Conversation 1. **Type a message** in the composer at the bottom of the screen. Describe what you need — for example: *"Summarise the key figures from this quarterly report"* or *"Draft an email to the supplier about the delayed shipment."* 2. **Hit send.** The agent begins working immediately. 3. **Watch it work.** The agent narrates what it's doing in real time - you'll see status messages describing each step, plus skill and tool cards appearing as they're invoked. 4. **Keep typing if you like.** You can send follow-up messages while the agent is still working - they queue up and are handled in turn - or stop the agent mid-task if it's heading the wrong way. ### Uploading Files You can upload files by: * **Dragging and dropping** files directly into the chat * **Using the attach button** in the composer Supported file types include: * **Documents** — PDF, Word (.docx), PowerPoint (.pptx) * **Spreadsheets** — CSV, Excel (.xlsx) * **Images** — PNG, JPG, and other common formats * **Email files** — EML * **Other text formats** — JSON, XML, plain text The agent processes uploaded files in a secure cloud sandbox. It can read PDFs visually (not just as extracted text), parse spreadsheets, and extract data from images. ### Understanding the Interface While the agent is working, you'll see several elements: * **Narration** - Short status messages describing what the agent is currently doing. * **Skill and tool cards** - When the agent uses a skill or an integration, a card appears showing its name and progress. Cards appear immediately in a loading state and update in place as they complete. * **Request cards** - When the agent needs something from you - an approval, an answer, or a file - a card appears in the chat for you to respond to. * **The right panel** - A side panel showing the agent's progress on longer tasks, generated files, the context it's using, and its memory. ### Tips for Your First Session * **Be specific.** The more context you give, the better the result. Instead of *"Process this file"*, try *"Extract all order line items from this PDF and create a CSV with columns for SKU, quantity, and unit price."* * **Upload files with your message.** The agent works best when it has the data it needs upfront. * **Follow up naturally.** You can ask the agent to refine, adjust, or build on its previous output within the same session. * **Check the narration.** If the agent seems stuck or headed in the wrong direction, the narration shows you what it's doing so you can redirect it. --- # 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 dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter, and the optional `goal` query parameter: ``` GET https://docs.lleverage.ai/operator-agent/getting-started.md?ask=&goal= ``` `ask` is the immediate question: it should be specific, self-contained, and written in natural language. `goal` is optional and describes the broader end goal you are ultimately trying to accomplish on behalf of the user. GitBook uses it to tailor the answer towards what is most useful for that goal. 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.