zotero-obsidian-bridge

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

• Third-party content exposure detected (high risk: 0.85). The required workflow reads Zotero item metadata/fulltext/annotations (including webpage items) via Zotero MCP or a local fallback, so outsider-authored paper or webpage text can be injected into the agent’s LLM context through the “read Zotero items from the target collection/query” → “get item full text / annotations / notes” path.