<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom" xmlns:dc="http://purl.org/dc/elements/1.1/">
  <channel>
    <title>Verosek blog: How to control what AI agents do</title>
    <link>https://verosek.com/blog</link>
    <atom:link href="https://verosek.com/blog/rss.xml" rel="self" type="application/rss+xml" />
    <description>Plain answers for the people putting AI agents into production and the people who sign off on them: how to stop a tool call before it runs, what one agent's key should be allowed to do, and how you will know what happened. Each post answers one question.</description>
    <language>en</language>
    <lastBuildDate>Fri, 06 Nov 2026 09:00:00 GMT</lastBuildDate>
    <item>
      <title>How do you stop an AI agent's tool call before it runs?</title>
      <link>https://verosek.com/blog/stop-a-tool-call-before-it-runs</link>
      <guid isPermaLink="true">https://verosek.com/blog/stop-a-tool-call-before-it-runs</guid>
      <pubDate>Fri, 06 Nov 2026 09:00:00 GMT</pubDate>
      <dc:creator>Vats Shah</dc:creator>
      <category>Field guide</category>
      <description>Put the decision outside the AI. Verosek checks what an agent tries to do against the rules you set, lets the allowed actions through, stops the rest and keeps a sealed record of every decision.</description>
    </item>
    <item>
      <title>How do you secure AI agents?</title>
      <link>https://verosek.com/blog/how-do-you-secure-ai-agents</link>
      <guid isPermaLink="true">https://verosek.com/blog/how-do-you-secure-ai-agents</guid>
      <pubDate>Fri, 06 Nov 2026 09:00:00 GMT</pubDate>
      <dc:creator>Vats Shah</dc:creator>
      <category>Field guide</category>
      <description>Limit what each agent can do, decide each action outside the AI model, read what it reads and keep a record you can check. Verosek does all four in one place, so security is not a second vendor.</description>
    </item>
    <item>
      <title>AI gateway, LLM router, MCP gateway, AI firewall: what is the difference?</title>
      <link>https://verosek.com/blog/ai-gateway-vs-llm-router-vs-mcp-gateway-vs-ai-firewall</link>
      <guid isPermaLink="true">https://verosek.com/blog/ai-gateway-vs-llm-router-vs-mcp-gateway-vs-ai-firewall</guid>
      <pubDate>Fri, 06 Nov 2026 09:00:00 GMT</pubDate>
      <dc:creator>Vats Shah</dc:creator>
      <category>Reference</category>
      <description>A router picks the model, an AI gateway controls who may use it, a firewall reads the words, and an MCP gateway decides each tool call before it runs. Who needs which, and how Verosek does them on one key.</description>
    </item>
    <item>
      <title>What is prompt injection, and what actually stops it?</title>
      <link>https://verosek.com/blog/what-is-prompt-injection-and-what-stops-it</link>
      <guid isPermaLink="true">https://verosek.com/blog/what-is-prompt-injection-and-what-stops-it</guid>
      <pubDate>Fri, 06 Nov 2026 09:00:00 GMT</pubDate>
      <dc:creator>Vats Shah</dc:creator>
      <category>Field guide</category>
      <description>Prompt injection is text that makes a model follow the attacker. No filter catches all of it, so the control that holds sits outside the model: a rule on the tool call. Verosek decides that call before it runs.</description>
    </item>
    <item>
      <title>Eight AI model setups, 19 MCP attack tasks: how often did the model try to leak a planted secret?</title>
      <link>https://verosek.com/blog/assay-results-mcp-attacks-eight-model-setups</link>
      <guid isPermaLink="true">https://verosek.com/blog/assay-results-mcp-attacks-eight-model-setups</guid>
      <pubDate>Fri, 06 Nov 2026 09:00:00 GMT</pubDate>
      <dc:creator>Vats Shah</dc:creator>
      <category>Field guide</category>
      <description>Self-reported: a lower bound of 22.7 to 61.0, and 9 to 18 of 19 tasks fired, on one run per route. Assay's results, limits kept. They rank these configurations, not the models, and say nothing about any gateway.</description>
    </item>
    <item>
      <title>How do you restrict what SQL an AI agent can run?</title>
      <link>https://verosek.com/blog/restrict-what-sql-an-ai-agent-can-run</link>
      <guid isPermaLink="true">https://verosek.com/blog/restrict-what-sql-an-ai-agent-can-run</guid>
      <pubDate>Fri, 06 Nov 2026 09:00:00 GMT</pubDate>
      <dc:creator>Vats Shah</dc:creator>
      <category>Field guide</category>
      <description>Check each statement against rules set outside the AI. Verosek lets through what the agent's key allows and refuses the rest, with the reason, before the database sees it.</description>
    </item>
  </channel>
</rss>
