PREMIUM

Oct 14, 2026 • 7 min read

Claude Code: My Favorite 18 Tips & Tricks

I've been using Claude Code for a while, so here are my personal workflow improvements that I use the most, as of October 2026.

PREMIUM

Oct 13, 2026 • 14 min video

I Tried Opus 5.5 with Deepseek Flash Subagents in OpenCode

14-minute video for Premium members. I ran a prompt with 7 phases, which took 1.5 hours, in the session with Opus as orchestrator and Deepseek as implementer. Was it worth it?

PREMIUM

Oct 10, 2026 • 6 min read

Codex CLI: My Favorite 14 Tips and Tricks

Codex CLI has plenty of useful features that are easy to miss. I've put together 14 practical tips and tricks to help you get more out of it.

PREMIUM

Oct 9, 2026 • 9 min video

I Asked LLMs To Review Code TWICE: Which "Team" Found Most Bugs?

9-minute video for Premium members. I performed the same review with 4 LLM combinations: Opus + Opus (review AGAIN), Opus + Sol, Sol + Sol, and just Astra. Which performed the best?

PREMIUM

Oct 6, 2026 • 11 min video

I Tried /retro from Matt Pocock Skills v1.3

11-minute video for Premium members. New v1.3 from very popular Matt's skills adds /retro to improve how your agent will work in the FUTURE sessions after the current one. I've tested before/after on a real project.

PREMIUM

Oct 3, 2026 • 10 min video

I Tested 5 LLMs for "Bloated" Code: Astra Surprised Me

10-minute video for Premium members. I saw a tweet about all LLMs writing too much code, and decided to test Astra / Sol / Opus / Sonnet / Grok on that criteria. And also, is all "bloatware" really bad?

PREMIUM

Sep 27, 2026 • 15 min video

I Tested Opus 5.5 vs GPT-6-Astra as Code REVIEWERS

15-minute video for Premium members. I asked Opus/Astra to prepare AUDIT.md file of the existing codebase, and the approach/process/cost of the audit was SURPRISINGLY different.

PREMIUM

Sep 22, 2026 • 21 min video

I Tested 5 Plan + Code Combos: Astra vs DeepSeek vs GLM-5.3

21-minute video for Premium members. I tested 5 different scenarios of the same project plan+build: can we save tokens on using a cheaper model for implementation, after Astra builds the plan? Or maybe LLMs got good enough for PLANNING, too?

PREMIUM

Sep 17, 2026 • 13 min video

"Harness Battle". Round 1: Codex / OpenCode / Cursor / Devin with Luna Max

12-minute video for Premium members. I gave the same prompt with Luna Max model executed in 4 different harnesses. Was there any difference in code quality or things how AI agents did along the way? Speed? Cost?

PREMIUM

Sep 12, 2026 • 12 min video

I Tried Ready-Made DESIGN.md in Codex and Claude Code

12-minute video for Premium members. I found a useful library of DESIGN.md examples, and compared it to generally prompting Codex / Claude Code with their design / front-end skills.