linear

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

• Third-party content exposure detected (medium risk: 0.30). Linear skill reads outsider-authored free text only from issues/comments that the user’s Linear team has labeled for the agent via the Linear MCP (e.g., “Read one task” get issue including comments, then “Record the outcome” create a comment), which an outsider could poison by posting into a monitored Linear issue/comment.