Sign in Book a demo
Back to Tools
Productivity & Docs OAuth
Build Your Agent View Power BI Docs
microsoft-powerbi logo

Microsoft Power BI MCP Tools for AI Agents

Let an agent inspect semantic models, query data with DAX, check refresh history, and start dataset refreshes in Power BI.

Available Tools

20 tools available in this integration

MicrosoftPowerbi.CheckDatasetGatewayHealth

Report whether each of a dataset's gateway-bound data sources is currently

MicrosoftPowerbi.DatasetGatewayHealth

Report whether each of a dataset's gateway-bound data sources is currently

MicrosoftPowerbi.GetDataset

Inspect one dataset in a single call: where it pulls data from, what parameters

MicrosoftPowerbi.GetDatasetRefreshHistory

Get the most recent refresh runs of a semantic model, with status and timing.

MicrosoftPowerbi.GetReport

Get a single Power BI report's detail, including the semantic model (dataset) it is

MicrosoftPowerbi.GetSchema

Get the full structure of a Power BI semantic model.

MicrosoftPowerbi.InspectDataset

Inspect one dataset in a single call: where it pulls data from, what parameters

MicrosoftPowerbi.ListDatasetRefreshes

Get the most recent refresh runs of a semantic model, with status and timing.

MicrosoftPowerbi.ListDatasets

List the Power BI semantic models (datasets) in a workspace.

MicrosoftPowerbi.ListGateways

List the data gateways the caller administers, with the id needed to reference

MicrosoftPowerbi.ListReports

List the Power BI reports in a workspace, each with the semantic model (dataset) it is

MicrosoftPowerbi.ListWorkspaceMembers

List who can access a workspace and the role each principal holds, for an

MicrosoftPowerbi.ListWorkspaces

List the Power BI workspaces (groups) the calling user can access.

MicrosoftPowerbi.QueryRows

Run a DAX query against a semantic model and return the rows as structured data.

MicrosoftPowerbi.RunDax

Run a DAX query against a semantic model and return the rows as structured data.

MicrosoftPowerbi.StartDatasetRefresh

Start an on-demand refresh of a semantic model.

MicrosoftPowerbi.TriggerDatasetRefresh

Start an on-demand refresh of a semantic model.

MicrosoftPowerbi.UpdateDatasetSettings

Repoint a dataset's parameters and/or change its refresh timing in one request.

MicrosoftPowerbi.ValidateDax

Check whether a DAX query is valid for a semantic model without returning its rows.

MicrosoftPowerbi.WhoAmI

Report the Microsoft account the toolkit is acting as: id, name, UPN, and tenant.

Why agent-optimized tools matter

Built for intent, not API schemas

Ask the agent to "inspect the sales dataset and explain why its last refresh failed." Arcade picks the right Power BI tools without guessing workspace, dataset, or report IDs.

Up to 100x fewer tokens

Power BI workspace and report reads return focused metadata, while the full semantic model schema is fetched only when needed, so the agent can choose the right dataset without loading every schema into context.

Complete the workflow end to end

Find the workspace, identify the report and semantic model, inspect its schema, validate and run DAX, and refresh the model through one toolkit designed around the complete Power BI task.

Make actions predictable and controlled

Arcade tools return structured results and useful errors across users and runs. Teams can apply scope restrictions, redaction, and approval policies so agents follow the same controls every time.

Securing Microsoft Power BI

The more capabilities agents get, the more useful they are. However, more capabilities bring more risks of undesired effects such as data loss and exfiltration.

This is why Arcade provides you with contextual rules that dictate how and when a tool is allowed to run. Arcade’s contextual access lets you govern tool visibility and behavior at every stage of execution and control who can see a tool, what inputs are allowed, and what comes back.

Learn about Contextual Access Policies

For Microsoft Power BI, you can implement some of these policies:

Access when tools are listed for a user
  • Hide MicrosoftPowerbi.TriggerDatasetRefresh from agents unless they operate under the data-platform operator role your IDP marks as authorized to start production refreshes.
  • Restrict MicrosoftPowerbi.UpdateDatasetSettings to administrators responsible for semantic-model configuration, so regular agents cannot change refresh timing or repoint production data sources.
  • Hide MicrosoftPowerbi.StartDatasetRefresh from agents that lack the data-platform operator role your IDP assigns to production refresh work.
Pre-execution before each tool call
  • Require human approval before MicrosoftPowerbi.TriggerDatasetRefresh runs against a production semantic model, so an agent cannot start an expensive or disruptive refresh without review.
  • Block MicrosoftPowerbi.UpdateDatasetSettings when it changes production connection parameters or refresh schedules outside your approved change window, and record the rejected request for audit.
  • Require human approval before MicrosoftPowerbi.StartDatasetRefresh runs on a production semantic model, so an agent cannot trigger a costly refresh on its own.
Post-execution after the tool returns
  • Redact connection strings, server names, database names, credentials, and parameter values from MicrosoftPowerbi.InspectDataset before they reach the model.
  • Strip email addresses, user principal names, and directory identifiers from MicrosoftPowerbi.ListWorkspaceMembers results while preserving each member's workspace role.
  • Redact the account's user principal name and tenant identifier from MicrosoftPowerbi.WhoAmI results so the model cannot expose personal or organizational identity details.
  • Redact hostnames, database names, and other connection details from MicrosoftPowerbi.CheckDatasetGatewayHealth results before they reach the model.
  • Strip connection endpoints, credential-like values, and private parameter contents from MicrosoftPowerbi.GetDataset results before the model can inspect them.
  • Redact customer names, email addresses, account numbers, and other direct identifiers from MicrosoftPowerbi.QueryRows results before the model sees the rows.

How to use Microsoft Power BI with your agent

Pick your MCP client below and follow the setup from authentication through your first Power BI request. Each end user authorizes their own Microsoft account.

Your questions, answered

A Microsoft Power BI MCP server exposes Power BI capabilities as structured tools that an AI agent can call through the Model Context Protocol. After a user authorizes their Microsoft account, the agent can inspect workspaces, reports, semantic models, schemas, refreshes, and query results, subject to that user's access.

Arcade provides an authenticated Power BI integration with per-user OAuth and an agent-optimized runtime rather than requiring you to build the authorization and tool-calling layer yourself. It also adds approval policies, response redaction, and controls for which actions an agent can access.

Arcade agents can list accessible workspaces, reports, datasets, gateways, and workspace members; inspect reports, datasets, schemas, gateway health, and refresh history; validate and run DAX queries; start dataset refreshes; and update dataset parameters or refresh schedules.

Arcade requires Microsoft Power BI permissions that match the actions you enable. Read-only work requires access to the relevant workspaces and semantic models, while refreshing models or changing dataset settings requires the corresponding Power BI permissions for those resources.

Yes. Each user authorizes their own Microsoft account through OAuth. Power BI actions run with that user's workspace, dataset, report, and tenant access rather than through one shared account.

Yes. Contextual Access policies can require human approval before consequential Power BI actions such as starting a refresh or changing dataset parameters or refresh schedules. The call proceeds only after the policy's conditions have been satisfied.

Arcade proxies Power BI API calls and does not persist your Power BI content as part of normal tool execution. Credentials are brokered and stored encrypted at rest, and the model never receives raw credentials.

Any framework or client that can connect to Arcade's MCP gateway or tool APIs works, including LangChain, OpenAI Agents, CrewAI, Claude Desktop, Claude Code, Cursor, Visual Studio Code, Microsoft Copilot Studio, and GitHub Copilot.

Arcade refreshes access while the user's Microsoft authorization remains valid. If it is revoked, rotated, or can no longer be refreshed, Power BI calls stop until the user reconnects the account.

Microsoft Power BI service limits, REST API throttling, tenant capacity, and workspace-level constraints still apply to calls made through Arcade. Refresh and query behavior can also be constrained by the capacity and permissions associated with the target semantic model.

Ready to actually ship your agent?

Connect your agent to Microsoft Power BI with managed OAuth, agent-optimized tools and approval controls for consequential actions.

Build Your Agent Read the Microsoft Power BI Quickstart