<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Upjectory Dispatches</title><description>Notes on growth, AI, and the modern martech stack from Alex Labbett.</description><link>https://www.upjectory.com/</link><language>en-us</language><item><title>Paperclip agents burn ~20,000 tokens before reading a task. Let&apos;s cut the bloat.</title><link>https://www.upjectory.com/blog/paperclip-token-optimization/</link><guid isPermaLink="true">https://www.upjectory.com/blog/paperclip-token-optimization/</guid><description>A hands-on playbook for cutting Paperclip&apos;s agent bill by over 80 percent without losing functionality. Baseline overrides, skill precedence, heartbeat timing.</description><pubDate>Sat, 18 Apr 2026 00:00:00 GMT</pubDate><category>AI</category><category>Agent Ops</category><category>Paperclip</category><category>Optimization</category></item><item><title>Welcome to Dispatches</title><link>https://www.upjectory.com/blog/welcome-to-dispatches/</link><guid isPermaLink="true">https://www.upjectory.com/blog/welcome-to-dispatches/</guid><description>Notes on growth, AI, and the modern martech stack. Written by hand from Los Angeles.</description><pubDate>Thu, 16 Apr 2026 00:00:00 GMT</pubDate><category>Martech</category><category>Growth</category><category>AI</category></item></channel></rss>