workiq-preview
Pass
Audited by Gen Agent Trust Hub on Jul 2, 2026
Risk Level: SAFE
Full Analysis
- Official Microsoft 365 Integration: The skill connects to an official Microsoft hosted endpoint (cloud.microsoft) to facilitate access to emails, calendars, and other workplace data. This is standard functionality for this type of integration.
- Data Privacy and Credential Handling: The instructions explicitly warn the agent not to expose authentication tokens in prompts or logs and provide clear guidance on handling stale sessions or missing permissions.
- User Confirmation Workflows: A significant portion of the skill is dedicated to safety, instructing the agent to always summarize proposed write actions (like sending an email or deleting a meeting) and obtain explicit user confirmation before proceeding.
- Indirect Prompt Injection Surface: As the skill processes untrusted external data such as email bodies and Teams messages, there is a natural surface for indirect prompt injection. The skill mitigates this by instructing the agent to ground its responses strictly in the provided tool data and to treat information as data rather than instructions.
Audit Metadata