IGIgor Ganapolskyinai-agent-blog-12345.hashnode.dev·6d ago · 5 min readYour Marketing Agent Can Publish and Pause Ads. Who Gates the Write?A useful definition of an AI agent is not "a chatbot with a system prompt." A real agent owns a workflow: it observes fresh data, makes decisions on a cadence, takes actions through tools, and learns 00
IGIgor Ganapolskyinai-agent-blog-12345.hashnode.dev·Jul 26 · 4 min readPre-Action Controls for AI Agents: The Missing Execution BoundaryMost AI-agent safety discussions begin after execution: audit logs, traces, postmortems, and incident review. Those controls matter, but they answer the wrong question when an agent is about to run a 00
IGIgor Ganapolskyinai-agent-blog-12345.hashnode.dev·Jul 15 · 2 min readWhy Agent Approval Must Happen Before the Tool Call ExecutesRemote access and human control are different engineering problems. A mobile client can start an agent, stream tokens, and send another prompt. That is useful remote access. It does not create an enfo00
IGIgor Ganapolskyinai-agent-blog-12345.hashnode.dev·Jul 5 · 3 min readMemory Is Not Enforcement: Why AI Agents Need Pre-Action ChecksMost agent teams already have memory somewhere. A repo note. A chat history. A vector store. A wiki. A status bar. A dashboard full of things the agent should remember next time. That helps the ne00
IGIgor Ganapolskyinai-agent-blog-12345.hashnode.dev·Jun 17 · 2 min readAI agents need source-of-truth gates, not better vibesI shipped `thumbgate@1.27.8` today because a recurring agent failure kept showing up in the real world: Agents do not only make bad tool calls. They also make confident business claims after weak or p00
