readwise-reader

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

• Third-party content exposure detected (high risk: 0.80). The skill explicitly fetches and returns untrusted, user-generated web content (articles, video transcripts, PDFs, tweets) from the Readwise service when using docs create with a --url and docs list/--with-content (see "Fetching content & transcripts" in SKILL.md), and the README instructs agents to fetch and analyze that html_content for summarization/triage workflows that can drive follow-up actions (e.g., docs update/pull), so third-party content could materially influence agent behavior.