connect-ai-exchange
Pass
Audited by Gen Agent Trust Hub on Aug 17, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill is designed as a specialized extension for CData Connect AI to interface with Microsoft Exchange. All operations use standard SQL or stored procedures mediated by the CData connector.
- [SAFE]: No indicators of malicious intent, prompt injection, or credential theft were found. The skill properly instructs the agent to confirm with the user before performing any write operations (e.g., sending emails, moving items, or responding to calendar invites).
- [INDIRECT_PROMPT_INJECTION]: The skill defines a surface for processing untrusted external data (email and calendar content) which could theoretically contain malicious instructions.
- Ingestion points: Data enters the agent's context through tables like Messages and Events in references/msgraph.md and references/ews.md.
- Boundary markers: None explicitly defined in the prompt templates.
- Capability inventory: The skill includes tools for sending and forwarding emails (SendMail, ForwardMail) and modifying calendar items.
- Sanitization: No specific sanitization instructions are provided for processed text fields.
- Note: This finding is assessed as low risk because processing such data is the core function of an Exchange connector and the skill enforces human-in-the-loop confirmation for all actions.
Audit Metadata