https://ofeng.org/tags/agents/ Recent content in Agents on Feng's Notes Hugo en Sun, 23 Aug 2026 00:00:00 +0000 https://ofeng.org/posts/designing-a-secure-ai-agent/ Sun, 23 Aug 2026 00:00:00 +0000 https://ofeng.org/posts/designing-a-secure-ai-agent/ <p>The hard part of securing an AI agent is accepting a slightly uncomfortable premise: <strong>prompt injection is not a bug we can reliably filter away.</strong></p> <p>An agent reads natural language from multiple places—user requests, emails, web pages, PDFs, search results, tickets, and tool responses. The model cannot reliably tell which of those words are authoritative instructions and which are merely data. A malicious document can therefore try to redirect the agent into revealing data, sending a message, changing infrastructure, or making a purchase.</p>