Daily Reading List – September 2, 2026 (#859)

What a week (already) for new LLMs. Claude Fable 5.1, Meta’s Muse Spark 1.3, and Gemini 3.8 Flash. And it’s only Wednesday.

[blog] Introducing Gemini 3.8 Flash and 3.8 Flash Cyber. Can’t stop, won’t stop. This is the third new Flash model in six weeks, and it’s another leap forward. Not only is it up there with frontier models on coding and agentic tasks, it’s a fraction of the price. And on most every Google surface, like Antigravity.

[blog] From trajectories to lineage: building better agent telemetry. One session log is interesting, but fades in usefulness if you’ve got multi-agent systems or teams of people working towards a goal with AI. Now you need agent linage to figure out how you arrived at a destination.

[blog] Audit your Agent files. Your configurations rot over time. Skills endlessly compound in your AI coding tool. Addy encourages us to audit regularly and test what happens when we prune.

[blog] A2A Is the Task. This is one of the better deep-dives I’ve seen into A2A use cases. And how it compares to MCP for those use cases.

[blog] 7 AI Agent Skill Patterns Every Programmer Should Know. I like the categories call out here. You’re probably building or using skills across one or many of these.

[article] Maybe We Shouldn’t Be Reviewing All This Code. And maybe we’ve asked too much of the humble code review. Shift judgement left?

[blog] 4 engineering patterns behind the strongest AI Agents Challenge submissions. It’s a good time to be studying patterns. We pulled these four common patterns from thousands of agents built by developers.

[blog] Agent memory as a file format. Are we over-processing agent memory? This writer proposes agent memory as portable data consisting of Markdown files and optional SQLite vector index.

[blog] Running Code OSS on Cloud Run instances: What Works, What Breaks, and What I Learned. Very creative. Run a personal cloud IDE for a few bucks a month within a Cloud Run instance.

[blog] Why my AI agents needed a rivalry. You might want more than one model in the mix so you don’t end with a homogenous team with the same quirks.

Want to get this update sent to you every day? Subscribe to my RSS feed or subscribe via email below: