Purpose
This tool is an MCP server that enables AI assistants like Claude or ChatGPT to interact with Power BI Desktop files programmatically. It allows users to run queries, modify measures, and manage data models through natural language commands.

Security Assessment
Because the tool runs locally and reads/modifies local Power BI files, it inherently accesses sensitive business data and proprietary model structures. The project claims to collect zero telemetry and states that all data processing happens locally on the user's machine, with the caveat that the connected AI assistant may send conversational data to its respective provider (e.g., Anthropic, OpenAI). The automated code scan could not verify the underlying source code, meaning potential hardcoded secrets, hidden network requests, or embedded malicious behavior cannot be definitively ruled out by this audit. Given the highly sensitive nature of business intelligence data combined with the unaudited source code, the overall risk is rated as Medium.

Quality Assessment
The project appears to be well maintained, with repository activity reported as recent as today. It has garnered 216 GitHub stars, indicating a solid and growing level of community trust. However, there is a discrepancy in the licensing: the automated scan detected "NOASSERTION," while the documentation explicitly states it uses a Proprietary license permitting personal and commercial use. A major quality concern is the inability to verify the tool's source code directly. Unlike standard open-source repositories, this project appears to distribute pre-compiled binaries, which restricts transparent peer review.

Verdict
Use with caution: while actively maintained and highly rated, the lack of accessible source code for deep auditing means you must implicitly trust the author with your local Power BI data.