<?xml version="1.0" encoding="UTF-8"?>
<rss version="2.0" xmlns:atom="http://www.w3.org/2005/Atom">
<channel>
  <title>Pipe Blog</title>
  <link>https://pipe-lang.com/blog.html</link>
  <description>Updates, tutorials, and deep dives into the Semantic Pipeline Runtime.</description>
  <language>en</language>
  <lastBuildDate>Wed, 05 Aug 2026 09:49:03 GMT</lastBuildDate>
  <atom:link href="https://pipe-lang.com/feed.xml" rel="self" type="application/rss+xml"/>
  <item>
    <title>Pipe v0.7.0 — Parallel, Self-Healing, and Tested by Default</title>
    <link>https://pipe-lang.com/blog.html?post=-Oz0foLhKpdmIf_BIEry</link>
    <guid isPermaLink="true">https://pipe-lang.com/blog.html?post=-Oz0foLhKpdmIf_BIEry</guid>
    <pubDate>Sun, 02 Aug 2026 08:55:17 GMT</pubDate>
    <description>Today we're shipping Pipe v0.7.0** — the biggest release yet. This version turns Pipe from a promising prototype into a language you can take seriously: parallel pipelines that just work, AI that fixes your code at runtime, a zero-setup test framework, module versioning, and production tooling for C</description>
  </item>
  <item>
    <title>Pipe: KI ist kein Library-Call mehr</title>
    <link>https://pipe-lang.com/blog.html?post=-OylHldmI9SGiJQFc4ai</link>
    <guid isPermaLink="true">https://pipe-lang.com/blog.html?post=-OylHldmI9SGiJQFc4ai</guid>
    <pubDate>Thu, 30 Jul 2026 04:31:57 GMT</pubDate>
    <description>Wenn du heute KI in ein Programm einbaust, läuft es immer nach demselben Muster: SDK importieren, Client instanziieren, API-Key verwalten, Retry-Logik bauen, JSON parsen, Exceptions abfangen — und das alles, bevor überhaupt der erste Prompt gesendet wird. Das ist nicht ineffizient. Das ist einfach d</description>
  </item>
</channel>
</rss>