yichen-grok-consult

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

• Third-party content exposure detected (medium risk: 0.30). In search_x_with_grok, the required workflow performs X keyword searches via the official Grok CLI x_search and—only if quota exhaustion is explicit—falls back to FxTwitter/OpenCLI/xreach to read public X content, meaning outsider-authored post text from X is ingested at runtime by the Grok CLI/X readers before GPT synthesizes it.