dare

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

• Third-party content exposure detected (high risk: 0.90). The skill explicitly requires fetching and reading open web content as part of its core workflow (SKILL.md lists "Web pages fetched (total)" and research loop targets) and references/abstractions.md enumerates SOPs and a tactic that perform web-searching/web-fetch/web-read (web-discover, web-fetch, web-read, web-research), so the agent will ingest and act on untrusted public web pages that could carry indirect prompt injections.