Oneself Highest Authority

Wall Street
Joined December 2008
Ok this is insane. No direction. No tools. No web search. Just pure Opus 5.5 from its own "imagination". Editorial storyboard, I have never seen anything one shot this with such great design.
1
67
This took opus about an hour, and used like $15 in api credits. It looks much better natively, this was recording using a free version of cap, so the res is not accurate.
12
Plain English v1.2 is out. It helps Cursor, Claude Code, Codex, and Grok explain technical work clearly—without changing facts, certainty, commands, or names. New: reader-aware explanations, tests, and one-command install. github.com/Tradelord223/plai…
5
3
68
Install it with: npx skills add Tradelord223/plain-english -g It supports Cursor, Claude Code, Codex, and Grok. The repo includes behavior tests and CI checks so every platform copy stays synchronized.
13
The skill keeps anything you may need to find or verify—file names, commands, error messages, flags, and configuration keys. It simplifies the explanation around those details instead of rewriting them.
5
It also adapts to the reader. An engineer can keep terms like “idempotent.” A store owner gets the behavior: retrying the same payment request returns the first result instead of creating a second charge.
4
Most “plain language” prompts just swap technical words for simpler ones. That can change the meaning. Plain English v1.2 explains what the system actually does while preserving evidence, uncertainty, requirements, and exceptions.
2
32
Codex browser use is cracked at solving captchas
58
I love @Notion because it works the EXACT way you would expect it to work, every time. It’s so fluid.
1
2
61
Grok 4.6 nor Fable can be trusted in production code. Sol seems to be the only model worth using in prod for a real company.
55
Lol the claude app literally never works
37
Finding a woman that will love you for who you are, and not for who she wants you to be is so fucking rare man.
30
I've been using @OpenAI's Daybreak Blue to do some bug bounties and I set it on a goal and literally in two hours it ran my $200 plan from 100% usage to zero. I've now hit my limit and the goal stopped mid-task. RIP
1
56
@thsottiaux @OpenAIDevs There may be a glitch. I can DM token usuage as proof. Less than 50 million tokens drained my limit in just over 2 hours.
1
23
I gave Grok 4.6, Claude Fable, and GPT Codex the exact same design brief. Three frontier models. Three radically different futures. One clear winner. My pick: GPT. Full comparison in the replies
1
51
Full-project scroll comparison
1
1
11
Individual long captures — Grok / Fable / Codex
11
Watching funny videos on Youtube