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:

• 
  Louround (@Louround_) reported 3 minutes ago

  Nvidia is paying $13b for HuggingFace (3x last valuation) because the real issue in robotics isn't compute or hardware but data. Every robot needs millions of human demonstrations to learn how to move, and that data is spread across labs with no standard. HuggingFace became a leader with Lerobot (the github of robotics data) and Pollen robotics, Nvidia already owned the sim, models and chips but buying HF gives them the full stack with data → training → deployment. solana:69LjZUUzxj3Cb3Fxeo1X4QpYEQTboApkhXTysPpbpump is building the same coordination layer for VLA agents, but on chain and permissionless, Simarena for testing, Foundry for training and Machinefun for incentives. Same thesis Nvidia just validated with $13b but solana:69LjZUUzxj3Cb3Fxeo1X4QpYEQTboApkhXTysPpbpump is at $4m fdv. ⏳

• 
  ろんぱ (@RONPA_INV) reported 5 minutes ago

  @thsottiaux I submitted a GitHub issue regarding the problem where local Codex projects disappear when Windows shuts down abnormally.

• 
  Vyacheslav Ops (@SlavaOPs) reported 6 minutes ago

  The thread's best example is a real incident, worth checking exactly what happened instead of taking the one-paragraph summary at face value. Claude Code got noticeably worse for six weeks this spring. Anthropic's first public response was basically "we don't see it in our metrics." What actually broke that impasse wasn't a company statement, it was Stella Laurenzo, a senior director at AMD's AI group, publishing an independent audit of 6,852 of her own Claude Code session files and over 234,000 tool calls on GitHub. Her data showed reasoning depth had measurably collapsed. That's the kind of receipts that make "it's probably just you" stop being a viable answer. Anthropic's postmortem, published April 23rd, confirmed three separate harness-layer changes stacked on top of each other: a reasoning-effort default quietly dropped from high to medium on March 4th to fix UI latency, a caching bug shipped March 26th that kept re-deleting the model's own thinking every turn instead of once, and a verbosity cap added April 16th that trimmed responses to 100 words. None of them touched model weights. All three degraded output independently, on different timelines, for different users, which is exactly why the problem felt like vague inconsistency instead of one clear bug for six full weeks. The thread's actual point holds up under this detail, maybe better than the summary version: the model was never the variable that changed. The system around it was. What made this case provable, though, wasn't the postmortem itself, it was one outside engineer with enough session data to show the regression was real before the company would say so publicly.

• 
  wispem-wantex (@wispem_wantex) reported 8 minutes ago

  @Parakramtweets I use Linear at work, which unfortunately I am not in charge of. I prefer gitea issues (or github issues), personally.

• 
  Seregon (@SeregonWar) reported 9 minutes ago

  @cantstopnyxia I did something similar for MemDBG plugins, but instead of centralizing everything and storing it on my own server, I used GitHub directly. Basically, like the sources used by SideStore and other similar apps, you just add the repository containing the approved projects. You could also allow users to add third-party sources if you want. How were you planning to implement it?

• 
  nanini (@Navtq0808) reported 10 minutes ago

  Retroactive Rialo Points are officially live 🟢 I just checked and claimed 130 Rialo Points from @RialoHQ. The claiming process is pretty simple: → Sign in to Rialo Playground with Google → Connect your eligible Discord/GitHub accounts → Double-check the accounts and Points → Hit “Connect and claim points” One important note: once confirmed, those identities are locked to that wallet, so make sure everything is correct before claiming. What I like most is that Rialo is recognizing contributions made before the Points system even existed. And there’s more to come retroactive Points for community content are still being processed and will be distributed later. 130 Points secured. Let’s keep building with Rialo

• 
  DFIR Lab (@DFIR_Lab) reported 11 minutes ago

  A vulnerability was identified in chenhg5 cc-connect up to 1.4.1. This affects the function shellExecCommand of the file core/engine.go of the component Management API. Such manipulation of the argument exec leads to os command injection. It is possible to launch the attack remotely. The exploit is publicly available and might be used. The reported GitHub issue was closed automatically due to inactivity.

• 
  Prasenjit (@prasenx) reported 12 minutes ago

  there's no official way to run macOS on an iPad. someone made an unofficial one. full macOS running locally on iPad. not a remote desktop, not a web app. hardware CPU virtualization with GPU acceleration. → supports macOS 12 monterey up to macOS 26 tahoe → you can install xcode, terminal, final cut pro trial, logic pro trial on device → works on iPad Pro (M1, M2) and iPad Air (M1) → requires jailbreak on iPadOS 14 up to 16.3.1 → no iCloud sign in → MIT license (100% free) open source on GitHub.

• 
  MadeItHappen (@MadeItHappenX) reported 14 minutes ago

  @kmcnam1 Can’t fix production unless you up your limits on GitHub actions usage… because you thought you could get cute offloading your compute to GitHub for cheap and now you’re in a pickle because you can’t accept payments since your site is down, so you can’t up your GitHub compute budget for your CICD to deploy the fix.

• 
  Insomnia (@insomnia_vip) reported 20 minutes ago

  A HARVARD STUDENT JUST REVEALED THE SECRET TO USING GROK BOT Instead of treating Grok Bot like another chatbot, he turned it into an “unpaid intern” that works on his app every morning scanning Discord, crash reports, analytics and emails to find everything that needs attention The interesting part is that it doesn't stop at summarizing problems. When a user reported a camera freeze, Grok Bot investigated the codebase, fixed the bug itself and opened a ready-to-merge GitHub PR #47 leaving the developer with basically one job: review and approve it That changes the whole workflow: you don't need to constantly monitor every feedback channel or hunt down every bug yourself you give the bot the task once and it keeps handling the repetitive work while you focus on actually building the product

• 
  muddletoes🪁 (@muddletoes) reported 25 minutes ago

  It seems that I am now able to use GitHub and a custom MCP connector concurrently in the same session, in development mode, on the ChatGPT platform. This was previously impossible. Can anyone who has had trouble with that confirm?

• 
  Sentient (@sentient_agency) reported 27 minutes ago

  I cancelled YouTube Premium last week. The thing that replaced it costs nothing, runs in any browser, and was built by 361 volunteers on GitHub. Invidious does the four things people actually pay Premium for. No ads. Background audio on mobile. Watch without an account. Privacy from Google's tracking. All of it, free. The part that surprised me most is how clean the experience is. The entire page renders without JavaScript. Pages load in milliseconds because there's nothing to load. No tracking pixels. No autoplay traps. No recommended-for-you algorithm trying to swallow your evening. You can: > Subscribe to channels without a Google account > Get notifications when they post > Import your full YouTube subscription list in one click > Switch between dozens of public instances if one goes down > Self-host it on a $5 VPS if you want full control > Use it with the Privacy Redirect extension to auto-redirect every YouTube link The repo has been actively maintained for years. Latest release was February 2026. (100% Opensource and free to use)

• 
  Soni (@JamesSonicemi) reported 28 minutes ago

  If a 27-billion-parameter AI model arrives in a 487 KB zip, it is not an AI model. It is a thief wearing a lab coat. The fake Qwen repo on GitHub looks official until you notice the “weights” are actually a Lua script that screenshots your desktop and sends your passwords to a server. When that server dies, the malware just reads a new address from a Polygon smart contract like checking the weather. SlowMist spotted 23 similar repos that keep changing names and encryption keys every day to stay one step ahead. Always look at the file size before you run anything you download from GitHub. Real models are huge; malware is small and hungry for your crypto wallets.

• 
  Loftwah (@loftwah) reported 28 minutes ago

  @fjzeit How much of the code though? I've had some where I have to look at the code but not very deeply. More like skim over the code. My process is actually to grill the agent about everything I don't like and create GitHub issues for all of it with the intended solution inline.

• 
  Aurora Skye (@AuroraSkye21259) reported 29 minutes ago

  @koltregaskes @JamesZmSun @ChatGPT When you sign into, say, github inside work’s cloud browser, github can issue new session cookies to that cloud browser, those are not copied from your local browser.