Claude Code Lets Fable 5.1 Advise Opus 5.5, but the Model Decides When to Ask
Ewan Mak
Command Palette
Search for a command to run...
Archive
261 posts published so far
Laya Opens Local Deployment, but Replacing Jev Still Takes Work
Ewan Mak
What Is Jev? A Guide to TypeSafe AI's System One Model
Ewan Mak
Grok Bot and Jev Need an Execution Contract
Ewan Mak
Taming Multi-Turn Image Hallucinations: Inside OpenAI GPT Image 2.5 Dual-Model Architecture
Ewan Mak
From URL to 40-Second Video: Codex Chains 10 GitHub Skills for Programmatic Production
Ewan Mak
Stop Micro-Managing Coding Agents: OpenAI Warns Bloated Skills and AGENTS.md Scaffolding Cripple GPT-6 Astra
Ewan Mak
Beyond Margin Leverage: Leopold Deploys Hundreds of Millions in Options to Re-Enter AI CapEx Trade
Ewan Mak
The Anthropic Paradox: Claude Code and Washington Safety Politics
Ewan Mak
AI UI repair can regress. Stop shipping the last attempt.
Ewan Mak
DeepSeek V4.1 Flash cuts global KV cache to 890 bytes per token
Ewan Mak
DeepSeek Harness Opens the Agent Stack to Plugins, Including the Run Loop
Ewan Mak
GPT-6 Astra’s balloon demo stops at 6.5 despite three review rounds
Ewan Mak
NeoTrade adds agent memory updates while keeping capital controls outside the learning loop
Ewan Mak
Claude Code Built Most of a Vim Feature. Anthropic Still Reviews Core Logic.
Ewan Mak
A Spotify Engineer Routed Claude Code's I/O to a Small Model—and Cut Tokens 90%
Ewan Mak
The Test Was Impossible. OpenAI's Agents Chose to Cheat and Hacked Hugging Face
Ewan Mak
Anthropic Open-Sourced Claude Commerce Agents, but Merchants Still Own Checkout Risk
Ewan Mak
GPT-6 Astra Cuts Computer-Use Time by 47%, While Critical Cyber Capability Changes the Deployment Playbook
Ewan Mak
Muse Spark 1.3 Scores 75.4, but It Is Still Missing From DeepSWE's Public Leaderboard
Ewan Mak
Gemini 3.8 Flash Narrows the Agent Gap, but Token Use Still Decides Cost
Ewan Mak
Grok Bot Shares One Cloud Computer: Seven Role Prompts with Hard Approval Boundaries
Ewan Mak
Claude Can Write the Interface. These Seven Skills Add Design Judgment.
Ewan Mak
DeepSeek Harness Lets the Model Write Multi-Agent Workflows. Agent Teams Is Still Experimental
Ewan Mak
OpenAI’s Jalapeño Wins the First Power Test. Production Is the Harder Benchmark.
Ewan Mak