> For the complete documentation index, see [llms.txt](https://docs.nexthink.com/platform/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.nexthink.com/platform/technical-previews/workspace-action-execution.md). # Workspace action execution {% hint style="warning" %} This Technical Preview is made available to customers free of charge for their evaluation and feedback; in general availability, the functionalities of the preview may be subject to additional cost and/or licensing.\ As such, the Technical Preview, the documentation, and any updates are provided for limited evaluation only and on an ‘as-is’ and ‘as-available’ basis without warranty of any kind. {% endhint %} ## Background Nexthink Workspace is the AI-powered, conversational experience embedded in the Nexthink Infinity platform. IT teams use it to investigate digital-experience issues in natural language, asking about device performance, application reliability, user-experience scores, and more, and Workspace analyzes endpoint telemetry to surface insights, trends, and recommended next steps directly in the conversation. Until now, Workspace could tell you what is happening and what to do next. With action execution, Workspace can now help you act on it, running remediation and automation directly from the conversation, always under your explicit control. ## What action execution does Action execution lets Workspace propose and, once you confirm, run operational tasks on your environment as a natural extension of the conversation. Instead of switching to another part of Nexthink Infinity to launch an action manually, you can go from diagnosis to remediation in a single flow. Two types of executions are supported: * **Remote actions:** scripts that run locally on targeted devices to collect information or apply a fix (for example, gathering GPO startup impact, clearing a cache, or restarting a service). * **Workflows:** orchestrated, multi-step automations that can combine remote actions, integrations, and logic to resolve an issue end-to-end. In both cases, Workspace only ever proposes executions that already exist in your Infinity tenant: it does not create new automations on the fly. ## Customer benefits * Go from diagnosis to remediation in a single conversation flow, without switching to another part of Nexthink Infinity. * Keep full control with human-in-the-loop confirmation: nothing runs without your explicit approval. * Reuse your existing role-based permissions, with no additional configuration required to keep executions safe. ## Permissions and access control Action execution uses exactly the same permission model as the rest of Nexthink Infinity. There is no separate or elevated set of privileges for Workspace. You can only execute the remote actions and workflows that your user role already grants you access to. Workspace operates strictly within your existing role-based access control (RBAC): if you cannot run a given action or workflow directly in Infinity, you cannot run it through Workspace either. Workspace never expands your permissions, and it never executes anything on behalf of another user. This means administrators do not need to configure anything new to keep executions safe: the guardrails you have already defined in Infinity apply automatically inside Workspace. ## The user journey Workspace decides, based on your question, whether an action or workflow could help, and then walks through a consistent, confirmation-first sequence: * You make a request in natural language (for example, "check the GPO startup impact on our Windows devices in Switzerland"). * Workspace identifies a candidate action or workflow and determines what it requires to run, such as input parameters or a target scope of devices. * Workspace clarifies the missing context with you conversationally, gathering everything the execution needs before proposing anything. * Workspace performs a risk assessment based on the number of devices in scope and the content of the remote action script. Based on this assessment, Nexthink's AI determines the risk level: low risk (no banner is displayed), medium risk (a banner is displayed, as shown in the confirmation card below), or high risk (a banner is displayed and an additional confirmation is required). For workflows, a high-risk warning is always displayed. * Workspace makes a proposal. A confirmation card appears in the conversation summarizing exactly what will run: the action, the targets, and any parameters. Workspace never executes an action without asking you to confirm first. * You approve or decline. You must choose one of the two options: you cannot continue the conversation until you have either approved or declined the proposed execution. * Workspace runs the action and shows progress. Once you approve, an execution card tracks progress in real time. Click **Get summary** to see the results of the execution. ### **The confirmation card** Before anything runs, Workspace shows a summary card so you can review the proposed execution and give explicit approval. Until you choose **Run** or **Discard**, the conversation is paused.
### **The execution progress card** After you approve, Workspace displays a progress card for the running execution. The **Get summary** button gives you details about how the execution went.
## What Early Adopters should know ### **Confirmation is always required** Action execution is human-in-the-loop by design. Workspace proposes; you decide. Nothing runs against your environment without an explicit approval, and the confirmation card doubles as a clear record of exactly what you agreed to run. ### **The conversation is intentionally blocked at the decision point** While a proposal is pending, you cannot continue chatting until you approve or decline it. This is deliberate: it prevents an execution decision from being lost in the flow of conversation and ensures every action is a conscious choice. ### **Target counts are point-in-time** The device count shown on a proposal reflects the moment the card is generated. Fleet membership changes over time, so if you need an authoritative, up-to-the-minute count, run an Investigation before executing. *** ## Scope of the preview This Technical Preview is available in all regions. Feature behavior, supported action types, and UI may evolve during the preview based on your feedback. If you are interested in participating or enabling additional capabilities, reach out to your Nexthink Customer Success Manager. ## Providing feedback Your feedback shapes how Action Execution graduates to general availability. By participating in this Technical Preview, you agree to: * Share how useful the proposed actions and the confirmation flow are in real troubleshooting scenarios. * Flag anything that felt unclear, risky, or unexpectedly blocked, especially around permissions and the confirmation step. * Take part in the occasional feedback session during the preview period. The quickest route for day-to-day feedback and questions is your Nexthink Customer Success Manager. --- # 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.nexthink.com/platform/technical-previews/workspace-action-execution.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.