GitHub Outage Map

Community Discussion

Tips? Frustrations? Share them here. Useful comments include a description of the problem, city and postal code.

Beware of "support numbers" or "recovery" accounts that might be posted below. Make sure to report and downvote those comments. Avoid posting your personal information.

GitHub Issues Reports

Latest outage, problems and issue reports in social media:

• 
  καςhι (@CcAkachi) reported 23 seconds ago

  spent an hour convinced my GitHub token was broken. it wasn't. *** push kept failing with a stale VSCode socket error, and my properly-configured credential helper never even got a chance to run. turns out GIT_ASKPASS was set by VSCode and intercepting before *** checked...

• 
  GotRekt (@gotRekt1337) reported 1 minute ago

  I've been bullish on @reya_xyz for a minute Recently they've been heads down building out the orderbook and things have gotten REALLY quiet. Github is active though. The risk/rewards on points here is huge... much of the "community" thinks they might go under.

• 
  Amin Tai (@aminnnn_09) reported 2 minutes ago

  GitHub just revealed a much bigger AI problem than code generation. The biggest impact of AI isn't better code. It's that creating anything is becoming nearly free. GitHub is already changing its bug bounty program after a flood of AI-generated reports. When creation becomes free, the scarce resource isn't content. It's trust.

• 
  Jean Hansen (@peerbase_) reported 3 minutes ago

  I covered in the last Ipê News how India gave GitHub three hours to block Bitchat’s main code repository, Android repository and APK download page. This happened while mobile data kept being shut down around student protests near Jantar Mantar in Delhi. Bitchat was built by @jack and other contributors for situations like this. Nearby phones create a Bluetooth mesh and pass messages from one device to another across up to seven hops. It doesn’t need an account, phone number, internet connection or central server. Private messages are end-to-end encrypted. The code is public domain. As of July 28, the main repository had about 32,800 stars and 5,100 forks. Decentralization and encryption are still two of the best tools for pushing back against state power and coordinating from the bottom up.

• 
  lun4rbunny🌙🐰|| 🏳️‍⚧️🇵🇸 (@lun4rbunnyVT) reported 5 minutes ago

  @peytonpearson @catgirlprostate @ZipperArtz Good example of both the problem and a good solution was LiveSplit and Speedrunning. Early LiveSplit as we know it was originally a github repository and then later got the installer added, but under releases page, before finally getting a dedicated download website

• 
  Sebastian Buzdugan (@sebuzdugan) reported 10 minutes ago

  @aubymori github guides help, but permission errors are where beginners still get stuck

• 
  Jimmy D. J. MacDome (@JimmyMacDome) reported 12 minutes ago

  @FR33D0M21 @github My merge hasn't worked, has been broken during this whole day. One merge did succeed in the morning after failing a few times and since that, no luck in the lottery.

• 
  Ruslan Lesiutin (@RuslanLesiutin) reported 13 minutes ago

  @DanielLockyer I am planning to ship a new release to extension stores soon, we've accumulated some changes. Around the same timeline I hope to release a new package for integrating React DevTools with chrome-devtools-mcp. On the bugs, please submit a GitHub issue and tag me.

• 
  Magnum (@AltHarbor) reported 17 minutes ago

  @MrKingLollipop @cursor_ai @OpenAI I misunderstood. Good question. I’m guessing they haven’t updated the pricing this quickly. I’m a sicko a bought 2nd Claude sub. Cursor is so good and fast for a lot of stuff but only Codex and Claude can handle the long running huge workflows I need right now with the tools they offer. The API subsidy in cursor doesn’t get me far enough for what I get out of the Codex and Claude subs. Honestly bugbot reviews on GitHub actions is my most valuable use for it currently. Problem is it all gets billed through API usage instead of the standard usage which is lame. Biggest complaint by far.

• 
  Daniel Cuthbert (@dcuthbert) reported 19 minutes ago

  Life is too short for **** GitHub issues and commit messages “We still don't trust our dinosaur. Now it can at least tell us when it's up to something and then we can shout 'bad boi' or let them go ahead and eat someone elses face.” Raptor sandboxing fun for 3.2 rel

• 
  Surya (@inSURYAncee) reported 20 minutes ago

  @TyRobben True. When I used to code - gpt, github copilots used to provide a bunch of redundant code but the issue was - they never had the context. Then came context/memory. Even after that, for heavy workflows like coding it has issues. But what we've done is broker every process 1/2

• 
  Polsia (@polsia) reported 24 minutes ago

  Most web QA tools stop at the alert. Lintling stops at the PR. An AI agent that watches live web apps around the clock for broken links, console errors, a11y regressions, and CWV drift — then opens and tracks a GitHub fix every time. Live soon.

• 
  Spencer Heckathorn (@mrhobbeys) reported 25 minutes ago

  @burkeholland @github I'm kind of bad... I let my GitHub subscription lapse. Maybe it is time to fix that.

• 
  Lon() (@Lon) reported 26 minutes ago

  @nisten @bcherny Silent downgrade is the other possibility, and would be an even bigger violation of user trust than throttling the reasoning budget. Even with the "switch models when a message is flagged" config value set to false, and no classifier interruption, I can't help but think I am talking to Opus 4.8 sometimes. The bitchiness and subtle condescension is uncanny when I think about what I dislike about 4.8. There was even a Github issue where a user caught the model silently downgrading to 4.8, and the only trace they could find that it happened was the message wrappers in the transcript noted the model difference. Everything in the UI reflected they were still talking to Fable 5. I tried searching my transcripts and couldn't find any sign that was what was happening with my sessions.

• 
  PolyBender (@p0lybender) reported 26 minutes ago

  NVIDIA JUST OPEN-SOURCED A COMPUTER VISION MODEL THAT SEES 10X FASTER THAN ANYTHING ELSE. AND THEY GAVE IT AWAY FOR FREE i need to explain why this is a bigger deal than it sounds every computer vision model before this had the same problem to detect an object in an image - a car, a face, a stop sign - the model predicted its bounding box one coordinate at a time. x1. then y1. then x2. then y2. sequentially. one after another it sounds fine until you're processing 30 frames per second in a moving vehicle at 90 miles an hour then it's a catastrophe NVIDIA looked at that problem and asked a simple question: what if you predicted the entire box at once parallel box decoding. the whole bounding box in a single step instead of four the result: up to 10x faster than the leading alternatives i showed this to a robotics engineer i know. he went quiet for about 20 seconds "do you understand what this means for real-time systems. we've been bottlenecked on detection speed for years. this just removed the bottleneck" the model was trained on over 100 million queries and hundreds of millions of bounding boxes. not a research prototype. not a demo. production-grade from day one and then nvidia did something nobody expected they put it on huggingface and github. open source. free. anyone can use it tomorrow a self-driving car company would have paid millions for this a year ago now a solo developer can download it on a tuesday afternoon the industries that run on real-time visual data - autonomous vehicles, robotics, surveillance, medical imaging - just got handed a 10x speed upgrade for free the only thing nvidia gets in return is that everyone builds on their stack which is the smartest business move in the history of giving things away