<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Ai on</title><link>https://blog.jmpsec.com/tags/ai/</link><description>Recent content in Ai on</description><generator>Hugo -- gohugo.io</generator><language>en</language><copyright>JMP Security. From Andorra with ❤️</copyright><lastBuildDate>Thu, 08 Oct 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://blog.jmpsec.com/tags/ai/index.xml" rel="self" type="application/rss+xml"/><item><title>Introducing osctrl-mcp</title><link>https://blog.jmpsec.com/post/introducing-mcp/</link><pubDate>Thu, 08 Oct 2026 00:00:00 +0000</pubDate><guid>https://blog.jmpsec.com/post/introducing-mcp/</guid><description>
The `osctrl-mcp` component extends osctrl's capabilities by exposing its read surface over the Model Context Protocol (MCP), enabling AI clients such as Claude Code or ChatGPT to inspect nodes, the osquery schema, and distributed query results through a permission-checked interface — read-only by default, with write tools gated behind an explicit switch.</description></item></channel></rss>