ClawHub

Microsoft Graph Api

Security checks across malware telemetry and agentic risk

Overview

This Microsoft Graph skill is not malicious, but it should be reviewed because it enables broad authenticated access to sensitive Microsoft 365 data without clear safeguards for risky API requests.

Install only if you trust the publisher and intend to let an agent operate through your Microsoft Graph/Microsoft 365 permissions. Use least-privilege OAuth scopes, prefer listed read-only actions, require explicit confirmation for POST, PUT, PATCH, DELETE, permission, mailbox, file, user, group, or directory changes, and revoke the Membrane connection when it is no longer needed.

SkillSpector

By NVIDIA
Vulnerability Patterns
  • Data ExfiltrationExternal Transmission, Env Variable Harvesting, File System Enumeration
  • Trigger AbuseOverly Broad Trigger, Shadow Command Trigger, Keyword Baiting Trigger
  • Prompt InjectionInstruction Override, Hidden Instructions, Exfiltration Commands
  • Privilege EscalationExcessive Permissions, Sudo/Root Execution, Credential Access
  • Supply ChainUnpinned Dependencies, External Script Fetching, Obfuscated Code
Findings (2)

Vague Triggers

Medium
Confidence
88% confidence
Finding
The skill description says to use the skill whenever the user wants to interact with Microsoft Graph API data, which is broad enough to trigger on many requests involving sensitive Microsoft 365 resources. Over-broad activation increases the chance the agent will invoke a high-privilege networked skill unnecessarily, causing unintended data access or modification.

Missing User Warnings

Medium
Confidence
94% confidence
Finding
The proxy request section encourages direct requests to Microsoft Graph but does not warn that these requests may read, modify, delete, or expose sensitive Microsoft 365 data. In this context, Graph commonly touches email, files, contacts, calendars, permissions, and directory data, so missing safety warnings materially increases the risk of destructive or privacy-impacting actions.

VirusTotal

VirusTotal findings are pending for this skill version.

View on VirusTotal