Antonio F

No fear, no distractions.

[Antonio F @anfed]

@thsottiaux hey - just got this on my Mac. Codex just completed a task..

[Antonio F @anfed]

@steipete @ATT I’m on T-Mobile and no complains at all

[Antonio F @anfed]

@thsottiaux not sure if it's on your radar, but pasting ‘text’ from MS Word to Codex is buggy. It’s pasted as an object that Codex can’t read. I have to paste it to a third-party tool (e.g., CotEditor) and then paste it back.

[Antonio F @anfed]

@DanCanobbio He’s given up on people. He likely has more haters than followers, so at this point in his career, he’s maximizing money.

[Antonio F @anfed]

@nickbaumann_ Is Linux support for the Codex app in the works?

[Sam Altman @sama]

codex with the $20 plan is a really good deal

[Antonio F @anfed]

Here’s a random idea: create a /notes section that can be either visible (configurable) or invoked. Just a thought.

[Antonio F @anfed]

@trq212, @thsottiaux -
I usually have 4-5 sessions open on each Claude/Codex. Sometimes, I forget what happened in each session. I’ve started using ‘stickies’ on each desktop to mark status.
How do you manage your sessions? Do you have other strategies besides relying on memory?

[Antonio F @anfed]

@werevertumorro el problema sera luego cuando los bancos no quieran darte acceso a la banca en linea por no tener una linea en Mexico.

[Antonio F @anfed]

@trq212 🤚 🤚 Not using OpenClaw, not anything like this.
Pure Claude Code user who is now consuming his 20x plan waaaay faster than before.

[Antonio F @anfed]

@rohanvarma I’m currently doing this because I’ve run out of tokens for the week. It would be great if Codex had a Claude migration skill that allows me to select all my Claude skills and migrate them to Codex.

[BridgeMind @bridgemindai]

Cancelling worked.

Anthropic just acknowledged the Claude Code rate limit issue.

GitHub issue #41788.

Max plan users hitting 100% in 70 minutes after v2.1.89.

Thousands cancelled.

Thousands switched to Codex with GPT 5.4.

Now they're fixing it.

Your wallet is the only feedback AI companies listen to.

Keeping my eye on Claude Opus 4.6 usage today.

If limits are back to normal, I resubscribe.

Stay tuned.

[Antonio F @anfed]

@lydiahallie After posting this, I’ve moved to Opus with thinking . I’m almost reaching my weekly usage limit on Thursday, even though I haven’t been pushing hard on changes.

[Antonio F @anfed]

@lydiahallie What about this:

https://t.co/v5FFJe0ZBb

[Boris Cherny @bcherny]

@jamonholmgren A little counter-intuitive, but I recommend using Opus with thinking for everything if you can. Because it’s more intelligent, it reduces both cost and latency vs. not using thinking or not using Opus

[Antonio F @anfed]

@dambuildshit @bcherny @AnthropicAI I’m a 200-dollar user and I’m waiting for the limits to reset for the very first time ever….

[Antonio F @anfed]

@rohanvarma @dkundel Something to run overnight. Don’t care if it’s done in 30 mins or 4hrs

[Antonio F @anfed]

@dkundel Hey @dkundel, why did you add the “disable-model-invocation” feature to the Codex Claude plugin? I was trying to add it to my existing skills.

[Antonio F @anfed]

@lydiahallie @trq212 Not only limits , for me skills are useless right now. Agents just decide to ignore instructions

[Antonio F @anfed]

@bcherny @Ra1kshit @claudeai "I don't think there's anything to fix in the skill. The instructions are clear and fast. The model just didn't follow them.The skill is well-structured with clear bash commands. The problem isn't the skill itself: it has exact scripts to run."

[Antonio F @anfed]

@bcherny @Ra1kshit @claudeai hey @bcherny , also skills are unreliable this morning.
from Claude:
"This is a Claude execution discipline problem, not a skill authoring problem. The skill literally has copy-paste-ready bash in Step 1, and the model went off-script."