article-agent-context-research

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

• Third-party content exposure detected (high risk: 1.00). The workflow explicitly delegates to Grok (x_search) via npm run grok:context and instructs the agent to search and collect sources from the open web (official docs, GitHub, secondary sites) and save URLs, so the agent will fetch and read untrusted public third‑party content.