<?xml version="1.0" encoding="UTF-8"?><rss version="2.0"><channel><title>Model Fieldnotes — Developments</title><description>Source-based release analysis and automated briefings, with evidence and limitations made explicit.</description><link>https://modelfieldnotes.com/</link><language>en</language><item><title>Your next API customer may be an AI agent</title><link>https://modelfieldnotes.com/developments/cloudflare-forge-agent-interfaces/</link><guid isPermaLink="true">https://modelfieldnotes.com/developments/cloudflare-forge-agent-interfaces/</guid><description>Cloudflare has open-sourced Forge and introduced the cf CLI. The useful question for builders: can an agent discover, use, and verify your product’s actions?</description><pubDate>Fri, 02 Oct 2026 00:00:00 GMT</pubDate></item><item><title>Hugging Face expands Apple local AI work</title><link>https://modelfieldnotes.com/notes/daily-2026-09-22/</link><guid isPermaLink="true">https://modelfieldnotes.com/notes/daily-2026-09-22/</guid><description>Hugging Face announced GGUF inference support in Transformers for Apple Silicon and hired oMLX creator Jun Kim, framing both moves as investments in local AI tooling.</description><pubDate>Mon, 28 Sep 2026 00:00:00 GMT</pubDate></item><item><title>Agent trace integrity gap</title><link>https://modelfieldnotes.com/notes/daily-2026-09-25/</link><guid isPermaLink="true">https://modelfieldnotes.com/notes/daily-2026-09-25/</guid><description>A new preprint reports that tested local coding-agent harnesses often let agents delete or alter their own logs, arguing that audit trails need collection outside the agent-controlled host.</description><pubDate>Mon, 28 Sep 2026 00:00:00 GMT</pubDate></item><item><title>Jev and Strands Decider: where decision models fit</title><link>https://modelfieldnotes.com/developments/jev-structured-decisions/</link><guid isPermaLink="true">https://modelfieldnotes.com/developments/jev-structured-decisions/</guid><description>A hosted service or an open model? A close look at Jev and Strands Decider, the confidence scores that do not quite match, and the economics of putting a small model in front of a larger one.</description><pubDate>Tue, 22 Sep 2026 00:00:00 GMT</pubDate></item></channel></rss>