ops-investigate-alert

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

• Third-party content exposure detected (high risk: 0.80). The skill's Step 8 ("Check Recent Code Changes (if gh available)") explicitly instructs the agent to call GitHub APIs (gh api repos/... tags and compare) and read commit messages/authors — untrusted, user-generated content that the agent will interpret to form root-cause hypotheses and next actions, exposing it to indirect prompt injection.