@zhouruijin For agents the real number is cost per finished task, not per token. A cheaper model that needs three retries can cost more than one clean Opus run. Has anyone measured that on Replit's agent yet?
@BenjieMalinao The classic one: the login screen hides the UI, but the data still ships to the browser. The fix is checking the session on the server (or row-level security) before any data leaves. What's the first signed-out check on your list?
@enzo66lin Both safe, different instincts: Claude Code read "clean" as undo, Codex read it as fix. Codex's way is nicer until the broken line is part of the WIP. Did you try a vaguer prompt like "start over"? That's where reset --hard risk usually shows up.
@ethanwalkerman "Default" is the word that matters. Most users never touch the model picker, so this moves a lot of usage overnight. Reads like a margin call more than a quality call. Will power users switch back to Claude, or just not notice?
@LDT0545 With limits off the table it comes down to how you work. Claude Code for long multi-file refactors where you stay in the loop, Codex for handing off parallel tasks and PR reviews. Which kind of work fills most of your week?
@Deeen_Codes Feels like it follows the crowd. After a wave of limit complaints people switch, the other tool takes the load, and the same complaints show up there a few days later. Are you hitting limits on long agent runs or on lots of short prompts?
@LucasEnkrateia Skills that explain instead of build are underrated, and pairing the HTML view with the skill means the context travels with the command. Planning to share it as a pack? Superpowers and Matt Pocock's skills top our ranking right now. prompttoprod.app/best/agent-ski…
@pankajkumar_dev The cheap escape hatch is an open-source agent where you bring the model. Kilo Code and Cline both run free models or any API key at provider rates, so there's no plan limit to hit, just a bill to watch. Is it the limits or the price that's hurting more? prompttoprod.app/best/free-ai-c…
@Kedeodev A for a solo build, mostly because Supabase bundles Postgres, auth and storage in one free tier, so you ship faster. The catch: free projects pause after a week of inactivity. Are you picking for speed to launch or for control later?
@FingValue Ouch, $40 for barely-better glitches hurts. Visual bugs are where agents loop worst, since they can't see the screen. A screenshot per attempt plus a hard spend cap usually beats letting it run all night. Hope the game gets there for her 🙂
@sprinntt That cost spread is the real story. At $1.02 vs $13.54 for the same scene, you could run Sol ten times and keep the best one. Was Opus's lighting worth the extra 50 minutes, or would a second Sol pass have closed the gap?
@ekcheungAI Agree on testing with a real repo. Stars show reach, but the bigger gap is the cost floor: OpenCode is MIT, ships free models and works with keys you already have, while Claude Code needs a Pro plan. Which one handled your last messy repo best?
556 Followers 547 FollowingAI builder 🇫🇷
https://t.co/IjjpjkOzam → Google Ads copilot
https://t.co/kGQ3SWB7KM → Free tool to check if your Codex / Claude Code subscription is worth it
148 Followers 183 FollowingBuilding https://t.co/Vs2jBxgvpE - a trip planner you can trust. Try it out! be my alpha testers!
Solo founder | Software Engineer | Amazon Go, Allen Digital
270K Followers 12 FollowingWe find the best price errors and glitches. We post them, you buy them before they are fixed. Follow us and turn on notifications. 🔔 Links may be affiliated.
601K Followers 42 FollowingGreatest deals finder of all-time 🐐 Best ones go fast - turn on alerts. As an Amazon Associate, Walmart, eBay, Lowes, & more I earn from qualifying purchases.
4K Followers 3K Followinga curious human who loves tech and talks around it 👩🏻💻 •
• spreading : https://t.co/ZQ8eBmINKj
wip: https://t.co/cfO09uJSJS , https://t.co/YnZ1HoXNkC
200 Followers 293 FollowingProduct Manager | AI Builder 🚀
Ship better products faster with AI tools & vibe coding | SaaS growth • Building in public
DM for collabs
2K Followers 2K FollowingTechie → Founder
Connect with builders on X: https://t.co/WPzpd4cusL 🚀
Fixing my own procrastination at https://t.co/m9xeA7yreo
Prev: https://t.co/NpDTlQ8wFY 💻
2K Followers 1K FollowingI don’t gatekeep. 10,000+ hours coding with AI since 2022. Building products until I can support my family. Dad 4 boys under 8, https://t.co/PUzG78pzqn, https://t.co/IVCXA65Biy
556 Followers 547 FollowingAI builder 🇫🇷
https://t.co/IjjpjkOzam → Google Ads copilot
https://t.co/kGQ3SWB7KM → Free tool to check if your Codex / Claude Code subscription is worth it
858 Followers 596 FollowingBuilding apps 🛠️ Making observations nobody asked for 🙃
Sharing what I learn about shipping & finding users.
Built ScreenFuel.
Scroll Less. Move More. Link ↓
415 Followers 464 FollowingBuilding stuff with AI:
https://t.co/9vjrnufWYh see if ChatGPT cites you
https://t.co/dNlUDeUTdy X newsletter
AniBlench my AI anime blender benchmark😅
543 Followers 515 FollowingCloud Developer II | Solo Developer 🚀
Building, learning, and sharing the journey.
Build Collab01Space 🤝
https://t.co/WD1l9JvQcb
3K Followers 2K FollowingBuilding profitable AI SaaS in public
https://t.co/9N6OAcmsPr • AI search visibility
https://t.co/PZN309nKtB • Reddit & X customer leads
236 Followers 274 FollowingBuilding @besidka_ai, an open-source BYOK AI chat on Nuxt + Cloudflare Workers, MIT. Technical Architect, 9+ years in ecom. I post what I learn shipping it solo