ultra-research

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

• Third-party content exposure detected (high risk: 0.90). The skill's recursive research loop and researcher prompts explicitly instruct agents to "scrape, summarize, save to L5/sources/" and to use live-web tools (Codex with live search, Perplexity fact-verifier, Exa web search) to ingest public web sources, meaning untrusted third‑party content is read and used to drive fact-checking, iteration decisions, playbooks and memory sync (see "RECURSIVE EXPLORATION" researcher prompts, "Perplexity Fact‑Verifier" and "Codex Cross‑Model Channel" sections).