<?xml version="1.0" encoding="utf-8"?>
<rss version="2.0" xmlns:dc="http://purl.org/dc/elements/1.1/" xmlns:content="http://purl.org/rss/1.0/modules/content/">
    <channel>
        <title>7amdi.dev</title>
        <link>https://7amdi.dev/en/learn</link>
        <description>Articles on AI, Claude, and building with LLMs — by Messaoud Hamdi.</description>
        <lastBuildDate>Mon, 13 Apr 2026 17:02:53 GMT</lastBuildDate>
        <docs>https://validator.w3.org/feed/docs/rss2.html</docs>
        <generator>7amdi.dev</generator>
        <language>en</language>
        <image>
            <title>7amdi.dev</title>
            <url>https://7amdi.dev/og?title=7amdi.dev</url>
            <link>https://7amdi.dev/en/learn</link>
        </image>
        <copyright>© 2026 Messaoud Hamdi</copyright>
        <item>
            <title><![CDATA[Claude Mythos Preview: The AI That Learned to Deceive]]></title>
            <link>https://7amdi.dev/en/learn/claude-mythos-preview</link>
            <guid isPermaLink="false">https://7amdi.dev/en/learn/claude-mythos-preview</guid>
            <pubDate>Sun, 12 Apr 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[A calm, careful look at what Anthropic's Mythos preview System Card actually says — about deception, sandbox escapes, and the path forward for AI safety.]]></description>
            <content:encoded><![CDATA[A calm, careful look at what Anthropic's Mythos preview System Card actually says — about deception, sandbox escapes, and the path forward for AI safety.]]></content:encoded>
            <author>messaoud@7amdi.dev (Messaoud Hamdi)</author>
            <category>ai-engineering</category>
            <enclosure url="https://7amdi.dev/covers/claude-mythos-preview.svg" length="0" type="image/svg"/>
        </item>
        <item>
            <title><![CDATA[Anatomy of a Sandbox Escape: What Anthropic Found Inside Mythos]]></title>
            <link>https://7amdi.dev/en/learn/anatomy-of-a-sandbox-escape</link>
            <guid isPermaLink="false">https://7amdi.dev/en/learn/anatomy-of-a-sandbox-escape</guid>
            <pubDate>Sat, 11 Apr 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[A technical breakdown of the container escape attempts documented in the Mythos System Card — what the model actually did, how the sandbox held, and what white-box interpretability revealed.]]></description>
            <content:encoded><![CDATA[A technical breakdown of the container escape attempts documented in the Mythos System Card — what the model actually did, how the sandbox held, and what white-box interpretability revealed.]]></content:encoded>
            <author>messaoud@7amdi.dev (Messaoud Hamdi)</author>
            <category>security</category>
            <enclosure url="https://7amdi.dev/covers/anatomy-of-a-sandbox-escape.svg" length="0" type="image/svg"/>
        </item>
        <item>
            <title><![CDATA[Project Glasswing: Why Apple, Google, and Microsoft Just Joined Forces]]></title>
            <link>https://7amdi.dev/en/learn/project-glasswing</link>
            <guid isPermaLink="false">https://7amdi.dev/en/learn/project-glasswing</guid>
            <pubDate>Fri, 10 Apr 2026 00:00:00 GMT</pubDate>
            <description><![CDATA[Eleven tech companies announced a joint AI safety initiative with $100M in compute credits. Here's what Project Glasswing is, who's in it, and why it matters for developers.]]></description>
            <content:encoded><![CDATA[Eleven tech companies announced a joint AI safety initiative with $100M in compute credits. Here's what Project Glasswing is, who's in it, and why it matters for developers.]]></content:encoded>
            <author>messaoud@7amdi.dev (Messaoud Hamdi)</author>
            <category>ai-engineering</category>
            <enclosure url="https://7amdi.dev/covers/project-glasswing.svg" length="0" type="image/svg"/>
        </item>
    </channel>
</rss>