> For the complete documentation index, see [llms.txt](https://move-solutions.gitbook.io/untitled/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://move-solutions.gitbook.io/untitled/mymove-manual/move-intelligence.md).
# Move Intelligence
Move Intelligence is the **AI agent** of MyMove. It answers questions in natural language on the **monitoring data of the project** (devices, measurements, alarms), reads the **documents** uploaded to the project knowledge base, generates **charts** and runs **skills**: reusable instructions saved in the project and invoked with **/**. Answers, charts and drafts are kept in the project as **conversations** and **artifacts**.
The feature includes:
* **Conversations** with the agent, with the steps of its reasoning visible.
* **Skills**: your own and the public ones of the project.
* **Documents**: the knowledge base of the project, shared by all its users.
* **Artifacts**: the charts generated in each conversation.
* **AI Balance**: included monthly usage and extra usage of the organization.
{% hint style="info" %}
The agent works on **one project at a time** and in **read-only** mode: it cannot modify or delete monitoring data, configure devices or manage alarms and notifications. It reports the data of the project and **cannot assess the safety of the structure**. AI can make mistakes, check important information.
{% endhint %}
{% hint style="info" %}
Move Intelligence is enabled per project. When it is not part of your plan, the page shows the **Upgrade required** message with the **Talk to Sales** action.
{% endhint %}
***
## Prerequisites
### Permissions
Move Intelligence uses the permission model of MyMove. Every project user except **Viewers** can use the agent; the **Operator** regulates skills and documents, the **Admin** domain regulates extra usage.
The agent accesses the data with the **same permissions of the user**: it never returns information that the user could not see elsewhere in the project.
## Conversations
### Start a conversation
On the first access the page asks **What are we working on?** and proposes some starting prompts (for example *Show sensors offline*, *Show alarms from the last month*, *Create a graph of displacement over the last 7 days*). Click a prompt or write your own message and press **Send**. The conversation is saved automatically with a title you can change later.
### Read an answer
The agent identifies the project data needed by the question, retrieves them and answers **in the language of the question**. While it works, the **Thinking process** lists the steps followed; the list stays available in the message once the answer is complete.
When the question needs a processing not available as such in the data (a statistic, a custom aggregation, a transformation), the agent writes and runs the code in an isolated sandbox without asking for confirmation, and returns the result. If the processing fails, the message **Sorry, the analysis could not be completed** appears with a **Retry** action.
Long lists of devices show the first five rows with **Show all** to expand them.
### Ask for a chart
Ask for a chart in natural language, or let the agent propose one when it helps the answer. The chart appears in the message and is added to the **Artifacts** of the conversation.
{% hint style="info" %}
Charts are drawn from the **raw data of the project**. Values computed by the agent in a processing (for example a moving average or a custom statistic) are returned as text or table and cannot be charted.
{% endhint %}
### Manage conversations
Each conversation in the list shows its title and last edit; the ⋮ menu offers **Rename** and **Delete** (with confirmation). Use **Search conversations** to find one by title.
## Skills
A skill is a request you repeat often, saved once and recalled with the `/move-skill-name` command. It is a markdown file with instructions in natural language: a **name** (`/move-skill-name`, assigned by the system when the skill is saved and never modifiable), a **description** the agent reads to decide when the skill applies, and a definition made of **Scenario**, **Steps** and **Outputs**. When a skill runs, the agent executes its steps in the order they are written.
### Use a skill in a conversation
Type **/** in the composer: the **Skills** popup lists the skills available to you with their description. Select one, add the details of the request after the chip and press **Send**.
The agent can also apply a skill **on its own** when your question matches its description, without the / command.
### Skills page
Available with the **Viewer** permission. The page lists your private skills and the public skills of the project, with their description and creation date; click a skill to read its markdown definition. The ⋮ menu offers, according to your permissions, **Edit skill**, **Use in a conversation** and **Delete skill**. Users with **Operator permission** can also delete public skills created by other users.
The **Conversations** tab keeps the drafting chats of skills not yet saved (**Continue chat**, **Delete**).
### Create a skill
Available with the **Operator** permission. Click **Add a new skill**: skills are written by talking, not in a form. Describe what the skill should do and when it should be used; the agent asks for the missing details and tells you if a step is not feasible, then proposes the definition (**Scenario**, **Steps** and **Outputs**) for review.
When you accept it, the agent proposes to save the skill in the project and asks whether it should be **public** or **private**. The skill is saved as private by default and appears in the Skills page; a draft left in the chat is not invocable until it is saved.
### Edit a skill
A skill is modified only by talking with the agent. **Edit skill** in the Skills page reopens its conversation: ask for the change, review the proposed **diff** of the definition and accept it. The name and the `/move-skill-name` command never change.
## Documents
The **Documents** page holds the knowledge base of the project: drawings, contracts, inspection reports and thresholds uploaded here become readable context for every user of the project, and the answers cite the page they come from. Each project has its own knowledge base, isolated from the other projects.
* With **Viewer** you can read the list of documents.
* With **Operator** click **Import new document** and drag and drop the files, or browse them. Each document shows its indexing status; once ready it is used in the following answers of every user.
* Deleting a document (⋮ menu, with confirmation) removes it immediately from the knowledge base: the following answers no longer use it. Answers already given are not changed and no user is notified.
## Artifacts
Every chart generated in a conversation is an **artifact** of that conversation. Click the **Artifacts** toggle in the conversation header (the badge shows their number) to open the panel: each row shows the chart title and creation date, and the comment icon (**Go to message**) scrolls to the message that generated it, so you never have to search the whole chat.
Click a row to open the artifact **full screen**. The actions bar offers **Go to message**, **Chart settings** and **Export**.
## AI Balance and extra usage
Every user (except Viewers) has a **monthly amount of tokens included in the seat**; the counter resets at the end of the month. Click **AI Balance** in the left panel to open the summary: **Your usage this month** (percentage of the included amount, reset in *n* days, billing cycle) and, for users with a Billing permission, the **Extra usage** status of the organization (Enabled or Disabled, monthly spending limit and extra usage of the month).
**Extra usage** lets the users of the organization keep using Move Intelligence after their included tokens run out; the extra consumption is billed to the organization at the end of the month. It is managed at organization level by users with **Admin permission:**
{% stepper %}
{% step %}
#### Open AI Balance settings
Click **Go to AI Balance settings** and confirm **Leaving the project?**: the conversation is saved and the page **Projects → Account → AI Balance** opens.
{% endstep %}
{% step %}
#### Enable extra usage
Turn on **Enable extra usage** and, optionally, set a **Monthly spending limit** in euro (empty = no limit). Settings are saved immediately.
{% endstep %}
{% step %}
#### Limit notifications
Billing users are notified when the organization reaches **80%** and **100%** of the limit.
{% endstep %}
{% endstepper %}
When the included tokens of a user run out and extra usage is off, or when the organization reaches its spending limit, **new messages cannot be sent**: the composer is disabled and a message explains the reason and that requests resume the following month. The agent resumes as soon as balance is available again.
## Troubleshooting
The page shows "Upgrade required"
Move Intelligence is not enabled for this project. Contact Move Solutions to add it to your plan.
The Skills or Documents page is not accessible
Your role has no **VIEW + Knowledge** permission. Ask a project administrator.
A skill does not appear in the / popup
The skill is private to another user, or it is still a draft in the **Conversations** tab: save it first.
The agent does not use a document
The document is still indexing or its indexing failed. Check its status in the Documents page and upload it again if needed.
The composer is disabled
Your included tokens for the month, or the spending limit of the organization, are exhausted. Users with the EDIT + Billing permission can enable extra usage or raise the limit in Account → AI Balance; otherwise requests resume the following month.
---
# 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://move-solutions.gitbook.io/untitled/mymove-manual/move-intelligence.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.
