Microsoft 365 MCP Server

• Analyzed: May 12, 2026, 3:05 AM
• Findings: 1
• Version: 1.0.0
• Hash: ddfd860708c074b4070db67562bf6f8f800d9b208f206cb6a7ad5cbd0fe8292b
• Source: palm
• External report: View on VirusTotal

Findings (1)

The skill is classified as suspicious due to the extremely broad Microsoft Graph API permissions it requests and utilizes, such as `Files.Read.All`, `Mail.Read.All`, `Chat.Read.All`, and `User.Read.All`. While these permissions are explicitly documented in `SKILL.md` and `README.md` as requirements for 'full Microsoft 365 integration' and the `src/index.ts` code only interacts with legitimate Microsoft endpoints (Azure AD for authentication, Microsoft Graph for API calls), the inherent capability to access and potentially exfiltrate a vast amount of sensitive organizational data (all files, emails, chats, and user profiles across the organization) represents a significant risk if the skill or the underlying Azure application credentials were ever compromised or misused. There is no evidence of intentional malicious behavior in the code itself, but the broad permissions elevate it beyond benign.