servicedesk-runbooks

HIGH W007: Insecure credential handling detected in skill instructions.

• Insecure credential handling detected (high risk: 0.90). The runbook explicitly shows embedding a plaintext temporary password in the Graph API PATCH body ("password": "{secureRandomPassword}"), which would require the LLM to generate or insert a secret value verbatim into requests/output, creating an exfiltration risk.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.65). The required workflow includes an approval gate where an approver can submit free-text comments (approverComments / approverResponse) via Teams Adaptive Card or approval connectors, and that free text is then stored/used in subsequent LLM context (e.g., logged/audited and potentially included in downstream prompts), making it outsider-authored free text at runtime.