Hitmanblood

Youtuber, PC Enthusiast, Gamer, Streamer

[Hitmanblood @Hitmanbloodtv]

@thsottiaux @buildwithrajath Keep being great Tibo! Dont forget the Sunday into Monday funday resets haha thanks again.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux Tibo. Truly. Thanks for resetting all this subsidized usage. Let’s people like me play around a little more with usage. And the bonus is we may or may not get a reset but it’s a great surprise when we do.

[Hitmanblood @Hitmanbloodtv]

@riotgames @Play2XKO Bro I never even knew you guys released this fighting game at all. Failure of advertising there

[Hitmanblood @Hitmanbloodtv]

@thsottiaux Same for features: I learned about larger context, /quit, imports, and the Profile dashboard through tweets, trial and error, or other AIs. Give Codex a built-in “What’s new / what can I do?” page sourced from the running version.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux Make Codex/ChatGPT explain limits. One page: exact limiter hit, requests, fresh/cache input, output, context, resets. /usage said 7.45B while deduped local telemetry was 11.55B, and chat timed out with 91% left. I shouldn't need Claude + 455 files to decode a $200 plan.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux BROOOOO

[Hitmanblood @Hitmanbloodtv]

@thsottiaux I’m a heavy ChatGPT Pro user and the last ~48 hours have been kind of insane. Not rage bait, I genuinely love the product, but this feels worth surfacing.

1/5

Today Health Home started throwing:

“MCP calls are not available in shared conversations”

Same account, same error on Chrome, a different browser, fresh Incognito login, AND iOS.

Meanwhile ChatGPT can still successfully access my connected Health records from a normal conversation.

So the data connection works. The Health Home UI doesn’t.

2/5

Then the iOS composer partially broke. Attachment button greyed out, then I couldn’t send the message at all. Had to move back to desktop.

Then my long-running project chat started returning:

“Request failed with status 429”

Fair enough, I use ChatGPT A LOT.

Until I opened a brand-new chat, typed literally:

“wassup bro”

...and immediately got the same 429.

3/5

Then I found other paying users reporting the exact same 429 / temporary conversation restriction at the same time, including a Plus user saying every model was unavailable and they were told to wait ~10 hours.

Status page: fully operational.

That makes it really hard as a user to know whether I hit a real limit, an account safeguard, or a service-side problem.

4/5

And yesterday I was trying to understand Sol’s context limitations versus other frontier products.

Today I find your post explaining that Sol can reason/work across multiple context windows using canonical compaction.

That’s a pretty important capability!

Why am I learning that from a random X post instead of ChatGPT clearly explaining what context/harness capabilities I actually have?

5/5

I’m not asking for unlimited compute. I’m asking for transparency.

Tell me:
• what limit I hit
• when it resets
• whether it’s account-level or service-level
• what context capabilities the product actually exposes
• when a feature is degraded even if the aggregate status page is green
Heavy users are exactly the people most likely to find this stuff.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux save me, i got 6 till 7. Reset time?

[Hitmanblood @Hitmanbloodtv]

@thsottiaux Lord member of the technical resets

[Hitmanblood @Hitmanbloodtv]

@thsottiaux reset time. I checked my calender it said at 11:42pm EST you guys would reset, we are almost at 11:43! you guys are late.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux Hey amazing weekend as a new codex user, man i have been having so much fun, one request, promise i wont ask for another XD. Can we get one more reset as a weekend success for how amazing Luna has been. Please and Thank you.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux Not sure if this is already in the works, but it’d be great to have an Auto mode in Codex that picks the right model + thinking level based on the task. That way simpler requests don’t burn higher-tier models and usage for no reason.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux Not sure if this is already in the works, but it’d be great to have an Auto mode in Codex that picks the right model + thinking level based on the task. That way simpler requests don’t burn higher-tier models and usage for no reason.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux you all are awesome! thanks.

[Hitmanblood @Hitmanbloodtv]

@thsottiaux Codex is helping me as a newbie "vibe code" things i always wanted as a kid and a gamer, im making dashboards for games i play and custom software! I don't have to hope and wait someone releases what i want! I can make anything now!

[Hitmanblood @Hitmanbloodtv]

@WOWCare, you guys should be embarrassed. Shit happens, lines get cut. Downtime happens. But your lack of communication sucks! You give us shitty, unrealistic deadlines. Then don’t give us updates. When I complain, I maybe get a $3 refund on my bill. Give a shit a little.

[Hitmanblood @Hitmanbloodtv]

Two things. There is no severe weather until around 7 or 8 p.m. Eastern tonight, so that’s fucking bullshit. When you posted this, it was dry as a bone and stayed dry until 4:30. It’s now 6:40 p.m., over two hours past your previous deadline, and you guys don’t give us a fucking update.

[Resident Evil @ResidentEvil]

𝙽𝚘 𝚜𝚠𝚎𝚊𝚝.

From the mind of Zach Cregger, director of Weapons and Barbarian. Watch the official trailer for Resident Evil. Filmed for IMAX – in theatres 9.18.

[Hitmanblood @Hitmanbloodtv]

@XBOX Hey guys you have a typo in your article. It’s a Radeon RX 6800s which is a mobile chip compared to the desktop 1070ti. It’s not a Radeon RX 68005

[Hitmanblood @Hitmanbloodtv]

This guy goes over the SCAM that is ashes of creation. How Steven tried to buy his guild back in the ArcheAge days. He has discord receipts and screenshots

https://t.co/LKIAQbAOlc