I love how Codex actually finishes what it was doing instead of just cutting you off mid-task when you hit your usage limits.
It can hit 0% usage left and keep working for 1+ hours.
Anyone can vibe-code a demo.
Show us the receipts.
— a wall of real projects shipped with GSD. Submit yours, top 3 win a year of GSD Cloud.
$GSD
@official_taches@gsd_foundation
Delete all of your skills, MCPs, plugins and CLAUDE/AGENTS.md.
Then start again.
Reinstall only what you know you need and can't live without.
Can't tell you how much better my coding agents are after removing 18 months of eclectic (and often conflicting) context buildup.
The Codex app really IS exceptional. So, so, so good.
That being said - it uses system resources like a motherfucker and I regularly have to quit it because it's launched 50+ Codex and Chrome for Testing helpers.
BREAKING: @BagsHackathon SUBMISSIONS ARE OFFICIALLY UNDER REVIEW 🔥
"1,444 projects applied, 6.9% of projects will win. The next generation of builders are on @BagsApp."
You're gonna love the new GSD Cloud "persistent planing" mode.
In this mode, your agent maintains awareness of the big picture goal across context clears.
You can start a new chat and it knows exactly what's done and what's up next.
You can also give it extra context that persists only for the duration of time it takes to complete a particular plan.
When the plan is complete that context is removed from future interactions with the agent.
Once the agent has written the plan, you can easily edit it yourself in the sidebar before execution.
You can even request specific human gate points where it will not continue until you're happy with its output.
So excited to share this with you guys soon.
It's @BagsApp@BagsHackathon demo day so here's a little preview of the soon to be released $GSD Cloud app I'm building for @gsd_foundation.
Manage AI agents on your home computer from ANY device with internet access.
Lots of very cool new features to share in the coming 2 weeks.