> For the complete documentation index, see [llms.txt](https://docs.decisionrules.io/doc/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.decisionrules.io/doc/ai-assistant/about-ai-assistant.md). # About AI Assistant ## About AI Assistant DecisionRules AI Assistant brings natural-language guidance and AI-assisted authoring directly into the DecisionRules application. It combines general language-model capabilities with DecisionRules-specific knowledge and the context of the page where you are working. The Assistant can help you: * Answer product questions using DecisionRules Documentation and Academy content * Find pages and controls in the application * Create Decision Tables, Scripting Rules, Lookup Tables, and Decision Flows * Prepare reviewable changes to existing Decision Tables, Scripting Rules, and Decision Flows * Plan and generate a complete process containing multiple rules and Decision Flows * Generate function expressions, test inputs, test suites, and rule summaries * Find a suitable template for a described use case Specialized agents handle these tasks. The agents use the current page, current rule, selected Decision Table cells, attached files, and other available context when the task requires it. {% hint style="info" %} Decision Flow Architect and Process Architect require DecisionRules App **1.26.2 or later** and AI Engine **1.3.0 or later** in self-hosted deployments. {% endhint %} ### Suggestions and Follow-ups When you open a new AI Assistant chat, predefined suggestions provide examples relevant to the current page and selected agent. Selecting a suggestion fills the input with a ready-to-use prompt. After some responses, the Assistant may also show suggested follow-up prompts. Select one to continue the same conversation and build on the current result. Suggestions are optional. You can always enter your own request instead. ### Agents Each specialized agent is designed for a particular task. You usually do not need to select an agent manually because the Assistant can route a clear request to the relevant capability. If you want to control the task directly, select an agent from the dropdown in the chat input. The available agents change according to the current page. For example, creation agents are available on the Rules List, while editing agents are shown on the corresponding rule detail page. See [AI Assistant Features](/doc/ai-assistant/ai-assistant-features.md) for the complete feature list and availability. {% hint style="success" %} Our team is continually extending and improving AI Assistant. If you are missing a useful feature or agent, contact our Sales team to discuss your use case. {% endhint %} ### Review Before You Apply AI-generated rule and process content is presented for review before it changes your space. * Creating a rule requires the explicit **Import Rule** action. * Editing a rule requires **Apply Changes**, followed by the normal **Save** action in the editor. * Generating a process keeps all generated content as a proposal until **Import process** is selected. Always inspect and test generated logic before saving or publishing it. ### AI Assistant Chat History AI Assistant saves conversations automatically. You can return to a previous session after closing the panel or navigating elsewhere. #### Where to Find It Open the AI Assistant panel and select **History** in the header.
AI Assistant History with saved, searchable conversations

Return to an earlier AI Assistant conversation from History.

#### What You Can Do You can: * Open a previous session and continue the conversation * Search loaded sessions by words from prompts, responses, titles, or rule names * Pin important sessions * Delete sessions you no longer need Sessions are ordered from newest to oldest and show their title, latest-message preview, and last updated time. When you switch spaces, the history refreshes to show sessions belonging to the selected space. ### Context-Aware Assistance AI Assistant can use more than the text entered in the current prompt. Depending on the selected agent and model, its context can include: * The current page or open rule * Selected Decision Table cells and current Test Bench data * Supported files attached with the attachment button, drag and drop, or clipboard paste * Shared Space Knowledge associated with the current space * Previous messages in the current chat session This context helps the Assistant respond to targeted requests without requiring you to repeat the complete background every time. ### Accessing AI Assistant AI Assistant is available on the **Lite Plan** or higher. It is also included in the **Free Trial**. Open it from the right panel by selecting **AI Assistant** in the upper-right corner of the page. Access also depends on the user's Space Role. The role must include the **Use AI Assistant** permission. Organization administrators can manage it under **Organization → Space Roles**. For Cloud and self-hosted configuration, see [Assistant Setup](/doc/ai-assistant/assistant-setup.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 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.decisionrules.io/doc/ai-assistant/about-ai-assistant.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.