The Agentic Engineer

I read the repos so you don't have to.

Weekly AI agent intelligence for builders. Every Wednesday.

Join … AI engineers getting smarter every Wednesday.

Check your email to confirm your subscription.

🔥

Deep Dives

One big story broken down with opinions. Not summaries — analysis.

⚡

Quick Hits

4-5 stories you need to know. Source links, star counts, HN points.

🛠️

Tool of the Week

Hands-on review with code snippets. Install it, try it, honest take.

📊

Agent Index

Weekly star tracker for the top 10 agent frameworks. Trends and moves.

📄

Paper Breakdown

One academic paper explained for practitioners. 5x compression.

🎧

Listen to Every Issue

AI-narrated audio on every issue. Hit play on the archive page or listen on the go.

🎙️

Podcast

Two-host audio version on YouTube, Spotify, and Apple Podcasts.

Latest Issues

Issue #21: GPT-5.6 Goes GA: Programmatic Tool Calling Changes Everything

GPT-5.6 ships three tiers and Programmatic Tool Calling. CubeSandbox gives your agents hardware-isolated sandboxes in 60ms. Anthropic finds Claude has a silent internal workspace.

Issue #20: Better Models, Worse Tools

Opus 4.8 invents fake schema fields and breaks Pi. Plus AWS CloudFormation Express Mode ships 4x faster deploys for agent-driven infra loops.

Issue #19: GPT-5.6 Sol: The First Government-Gated Model Release

OpenAI gates GPT-5.6 Sol behind government approval. GLM 5.2 beats Claude on cyber benchmarks. AWS Lambda MicroVMs give coding agents proper sandboxes.

Issue #18: Claude Requires Government ID: The Trust Fracture

Anthropic demands government ID for Claude access. 754 HN points of fury. AWS Continuum launches autonomous security agents. GLM-5.2 gives the open-model exodus a landing pad.

Issue #17: The First Frontier Model Recall

US government pulls Fable 5 and Mythos 5 from production. AWS ships three agentic services in one week. AWS Blocks ships a local-first backend toolkit with AI steering baked in.

View all issues →

Listen to the Podcast

YouTube Spotify Apple RSS