<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0"
  xmlns:atom="http://www.w3.org/2005/Atom"
  xmlns:content="http://purl.org/rss/1.0/modules/content/"
  xmlns:dc="http://purl.org/dc/elements/1.1/"
>
  <channel>
    <title>Memento Blog</title>
    <link>https://runmemento.com/blogs/</link>
    <atom:link href="https://runmemento.com/blogs/rss.xml" rel="self" type="application/rss+xml" />
    <description>Field notes from building Memento — a local-first memory layer for AI assistants. Opinion-led, technically honest, updated as the MCP ecosystem moves.</description>
    <language>en-us</language>
    <lastBuildDate>Sun, 17 May 2026 00:00:00 GMT</lastBuildDate>
    <item>
      <title><![CDATA[I tried every AI memory tool. Here's why I built another one.]]></title>
      <link>https://runmemento.com/blogs/i-tried-every-ai-memory-tool/</link>
      <guid isPermaLink="true">https://runmemento.com/blogs/i-tried-every-ai-memory-tool/</guid>
      <pubDate>Sun, 17 May 2026 00:00:00 GMT</pubDate>
      <dc:creator><![CDATA[Raghu]]></dc:creator>
      <description><![CDATA[AI tools call it 'memory' — but it's a flat vector store with no types, no audit, no decay, no conflict detection. Memento is memory built like infrastructure: typed, audited, decay-aware, local.]]></description>
      <category>launch</category>
      <category>opinion</category>
      <category>mcp</category>
    </item>
  </channel>
</rss>
