obsidian-knowledge

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly ingests and acts on external web content—e.g., "If the user gives a URL,正文,文章或报告...先委托 read" and the Knowledge Capture/Delegation sections specify using fetch/read to extract and save webpage content and URLs into the vault—so untrusted third-party pages can be read and used to drive create/update actions.