graphify

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

• Third-party content exposure detected (high risk: 0.90). This skill explicitly instructs ingesting content from public websites in the "Graph Extraction Sources" -> "From Websites" section (e.g., the html.querySelectorAll('a[href]') link-extraction example) and then uses that extracted, untrusted web content to build graphs, define graph-based tools, and drive agent orchestration, so arbitrary third‑party pages could materially influence agent behavior.