PinnedShri·Feb 25Claude Code vs OpenAI Codex: A Detailed ComparisonBoth Anthropic’s Claude Code and OpenAI’s Codex (GPT-5.3-codex) are advanced AI coding assistants, but they have different design…
Shri·Oct 10Claude Code vs Codex (8 Months Later): Which One I'd KeepEight months of real work with Claude Code and OpenAI Codex. Why I'd keep Codex, and the one Claude seat I'd still pay for.
Shri·Sep 20I Let GPT-6 Astra Build My Mobile App. It Felt Like Jarvis.SupWriter already makes money. Customers wanted a mobile app. I gave Astra the tools and kept the feedback loop running.
Shri·Sep 16I Handed My SEO, My Ads and My Blog to Three Bots That Never Log OffFour weeks on Grok Bot: the exact three-agent setup, the templates worth stealing, and the line in the docs nobody reads until it bites.
Shri·Sep 13GPT-6 Astra Is the Best Model I’ve Ever Used. It Writes Almost None of My Code.I dropped Claude Fable 5.1 for it in three days. Then the $200 plan taught me what Astra is actually worth paying for.
Shri·Aug 13I Made Claude My Tech Lead. It Runs Cursor and Codex as Subagents -While I Sleep.Fable plans, Grok codes, Fable reviews the diff cold, and Chrome click-tests the result. The exact setup, configs included.
Shri·Aug 11Talking to My Mac 10x’d My Dev Productivity: 700,000 Words in 6 MonthsDictating prompts to my AI agents is the cheapest multiplier I’ve found all year. Here’s why it works — and why the local version beats…
Shri·Jul 16GPT-5.6 Sol vs Claude Fable 5 : I Used Both for 7+ Days — Here Are the 7 Rules Nobody Tells YouSol isn’t smarter than Fable. It’s something more dangerous: a model that never stops. Here’s how I split my work between them now.
Shri·Jul 15I Used GPT-5.6 for 7 Days Straight. It’s a Workhorse, Not a Genius.The honest review after a week of real work: what blew me away, what’s still broken, and whether it’s worth switching from Claude.
Shri·Jul 7The real skill in AI coding is knowing what you don’t know.Working with Fable keeps teaching me the same old lesson: the map is not the territory.