<?xml version="1.0" encoding="utf-8" standalone="yes"?><rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom"><channel><title>Vylen — AI Command Center for User-Owned Hermes Agents</title><link>https://vylenagent.com/</link><description>Recent content on Vylen — AI Command Center for User-Owned Hermes Agents</description><generator>Hugo</generator><language>en-us</language><lastBuildDate>Fri, 22 May 2026 00:00:00 +0000</lastBuildDate><atom:link href="https://vylenagent.com/index.xml" rel="self" type="application/rss+xml"/><item><title>Same chat, every screen</title><link>https://vylenagent.com/blog/same-chat-every-screen/</link><pubDate>Fri, 22 May 2026 00:00:00 +0000</pubDate><guid>https://vylenagent.com/blog/same-chat-every-screen/</guid><description>&lt;p&gt;&lt;img src="https://vylenagent.com/img/blog/same-chat-every-screen-hero.webp" alt="Three Vylen clients open on the same chat at the same time: Linux Chrome, the Vylen iOS app on an iPhone, and Windows Chrome, each showing the same assistant message mid-stream."&gt;&lt;/p&gt;
&lt;p&gt;Earlier this week I had the same conversation open on three screens at once. Linux Chrome at my desk. Windows Chrome on a second machine. The Vylen iOS app on my iPhone, propped against my keyboard. I typed a question on the desktop and pressed enter.&lt;/p&gt;</description></item><item><title>Hermes Agent (GitHub): Self-Host Guide, Setup &amp; Daily-Use Notes</title><link>https://vylenagent.com/hermes-agent/</link><pubDate>Thu, 21 May 2026 00:00:00 +0000</pubDate><guid>https://vylenagent.com/hermes-agent/</guid><description>&lt;style&gt;
.hg-toc {
 margin: 24px 0 36px;
 padding: 18px 20px;
 border: 1px solid rgba(139, 92, 246, 0.35);
 border-radius: 12px;
 background: rgba(11, 16, 32, 0.6);
}
.hg-toc strong {
 display: block;
 margin-bottom: 10px;
 font: 700 12px/1.2 JetBrains Mono, ui-monospace, SFMono-Regular, Menlo, monospace;
 letter-spacing: 0.08em;
 text-transform: uppercase;
 color: #9aa6c8;
}
.hg-toc ol {
 margin: 0;
 padding-left: 20px;
 color: #c8d0e8;
}
.hg-toc ol li {
 margin: 4px 0;
}
.hg-toc a {
 color: #c8d0e8;
 text-decoration: none;
 border-bottom: 1px dashed rgba(139, 92, 246, 0.4);
}
.hg-toc a:hover {
 color: #fff;
 border-bottom-color: #8B5CF6;
}
.hg-glance {
 margin: 16px 0 28px;
 width: 100%;
 border-collapse: collapse;
}
.hg-glance th,
.hg-glance td {
 padding: 10px 12px;
 text-align: left;
 border-bottom: 1px solid rgba(104, 124, 255, 0.18);
 color: #dfe6ff;
 font-size: 15px;
}
.hg-glance th {
 width: 30%;
 color: #9aa6c8;
 font-weight: 600;
 font-size: 13px;
 letter-spacing: 0.04em;
 text-transform: uppercase;
}
.hg-callout {
 margin: 22px 0;
 padding: 16px 18px;
 border-left: 3px solid #8B5CF6;
 background: rgba(43, 25, 89, 0.32);
 border-radius: 0 10px 10px 0;
 color: #eef2ff;
}
.hg-callout strong {
 display: block;
 margin-bottom: 4px;
 color: #c4b5fd;
 font-size: 13px;
 letter-spacing: 0.05em;
 text-transform: uppercase;
}
&lt;/style&gt;
&lt;p&gt;&lt;strong&gt;Hermes Agent&lt;/strong&gt; is an open-source, self-improving AI agent built by &lt;a href="https://nousresearch.com"&gt;Nous Research&lt;/a&gt;, with source code on &lt;a href="https://github.com/NousResearch/hermes-agent"&gt;GitHub at NousResearch/hermes-agent&lt;/a&gt;. It runs locally or on a small VPS, uses any LLM provider you choose, executes real tools (terminal, code, file, browser), creates and improves its own skills over time, and talks to you from Telegram, Discord, Slack, WhatsApp, Signal, the CLI — or, with Vylen, from a native mobile app built for it.&lt;/p&gt;</description></item><item><title>The Scary Part: Hermes Approval Depends on Which Tool the Model Picks</title><link>https://vylenagent.com/blog/hermes-approval-gate-tool-choice/</link><pubDate>Wed, 20 May 2026 00:00:00 +0000</pubDate><guid>https://vylenagent.com/blog/hermes-approval-gate-tool-choice/</guid><description>&lt;p&gt;While testing Vylen&amp;rsquo;s new MessageEvent chat path, we hit a behavior that is easy to miss and hard to unsee:&lt;/p&gt;
&lt;blockquote&gt;
&lt;p&gt;In Hermes, whether a dangerous operation asks for approval can depend on which tool the model decides to use.&lt;/p&gt;
&lt;/blockquote&gt;
&lt;p&gt;That is not a Vylen rendering issue. Vylen showed the approval card when Hermes requested approval. The surprising part is that the same user intent later succeeded without an approval prompt because Hermes chose a different tool path.&lt;/p&gt;</description></item><item><title>Hello, Vylen</title><link>https://vylenagent.com/blog/hello-vylen/</link><pubDate>Tue, 12 May 2026 00:00:00 +0000</pubDate><guid>https://vylenagent.com/blog/hello-vylen/</guid><description>&lt;p&gt;Most AI tools today ask you to upload your context, your keys, and your work to someone else&amp;rsquo;s machine. Vylen takes a different bet: &lt;strong&gt;your agents should run on your hardware&lt;/strong&gt;, and the controller you use to drive them should be calm, fast, and yours.&lt;/p&gt;
&lt;p&gt;Vylen Agent is a control plane for &lt;a href="https://vylenagent.com/hermes-agent/"&gt;user-owned Hermes Agent&lt;/a&gt; instances. The job of the controller is small but specific: pair a connector, hold a single outbound tunnel open, route commands to your local agent, and give you a mobile-first surface to see what it&amp;rsquo;s doing. The Hermes-side gateway plugin is open source (&lt;a href="https://github.com/GiorgioRegni/hermes-vylen-gateway"&gt;hermes-vylen-gateway&lt;/a&gt;, MIT). The client app is a private project, public release pending.&lt;/p&gt;</description></item></channel></rss>