gitlab

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

• Third-party content exposure detected (high risk: 0.75). The skill’s runtime path uses glab mr view/diff/note list and glab ci trace to fetch MR notes and CI job logs from GitLab; those are outsider-authored free text (e.g., MR discussion/comments and build failure logs) that can be ingested into the agent’s LLM context for summarization.