atlassian-rest

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

• Third-party content exposure detected (high risk: 1.00). The skill explicitly fetches and reads user-generated content from Confluence pages and Jira issues (e.g., workflows like workflows/spec-to-backlog.md, workflows/search-company-knowledge.md, and workflows/capture-tasks-from-meeting-notes.md call confluence.mjs get-page and jira.mjs search/get) and then uses that content to drive actions such as creating epics/child tickets or synthesizing answers, so third-party page content can materially influence the agent's actions.