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:

• 
  Sherry Banks (@segetayoo3) reported 9 minutes ago

  @Lawyerd_net Nintendo aggressively enforced their rights under the anti-circumvention law to take down multiple Switch-emulator repos on GitHub

• 
  ꝠꭵꝇꝇꭵaꝳꟻꞨ (@SeadAwkward) reported 9 minutes ago

  @LardManSmith64 Nintendo is creating a challenge. They're pushing emulation into more low-profile places. Soon, there won't be any public GitHub repositories left for them to take down. There won't a "face" to sue.

• 
  umang (@umgbhalla) reported 11 minutes ago

  @troels_a make it beep when github is down

• 
  Zook (@zook_data) reported 12 minutes ago

  @Andy_ShuoYang @grok, study the OP linked Github and the arxiv paper. Explain the "bandwidth-adaptive CPU-GPU" feature and in what systems it truly shines. Tell me the limits or compatibility issues with Freetoken (e.g. in what systems/models it won't outperform). Finally, tell me if Freetoken can produce a OpenAi compatible endpoint I can plug and play into my existing coding agent harness.

• 
  Shaw (spirit/acc) (@shawmakesmagic) reported 16 minutes ago

  @FudaXiv @DrNickA You are a lying scammer, this is fake GitHub commits Blocking you, you are the problem with this space Absolutely retarded I get so much hate from retards because of ********* like you

• 
  Suhail Roushan (@0xsuhailroushan) reported 18 minutes ago

  cursor looked at github's outage and said “interesting” then built github

• 
  L👀ty McFreely (@Vote4YourLeader) reported 21 minutes ago

  @BarronNaledi Each pass gets easier. It remembers projects. I've trained it to learn how I work so that I don't have to start fresh with those instructions every time. I just give it an instruction file and tell my agent (Cassian - means empty in Latin, because I'm developing him for public github repo for anyone to use) the filename and it kicks off to my specification. Next time it kicks off I'll have developed it more with my other projects, so I'll get a better result. It does slow things down, I make it do a lot of work not just spit something out. It's infinitely worse for my video generating skill, that thing is burning unbelievable usage training. Research and software projects are locked in. Give spec, set to work, wait, iterate.

• 
  The big D (@siegmacat) reported 24 minutes ago

  @ll7HaloBeastll7 @SunriseDevTeam "They use AI" is a critique in itself, the one even acknowledged in github readme. But, of course, it's so convenient how you can play old D2 with increasingly more features that suddenly it's not a problem and moral side of the question doesn't matter

• 
  Patrick J. Auld (@PatrickAuld) reported 27 minutes ago

  I can confidently say local development is dead. I've run remote agents most of the year, and the only reason I run something locally is for a permissions issue. Local agents only found a foothold because it allowed people to grant agents greater access. I priced out my M2 MacBook against GitHub Codespaces when I got it. Codespaces was obviously cheaper. But startup latency was bad, and the tooling managing them wasn't very mature. Agents feel similar. There are options, but none make a ton of sense over running a box yourself. Running the model locally or calling a provider is orthogonal. You want an always-on version of your agents that's accessible across other devices The AI pin or the screenless interface is making more sense. A substantial amount of work can be done if we figure out remote agents with appropriate permissions beyond what the company that runs them can self-provide. Finding the balance between persistent, ephemeral, swarms, and all the complicated parts made simple will be the unlock. Their infinite patterns right now. They're all sort of good at different verticals. The reason all these exist is basically just context management. An agent with sufficient context to run for a long time and have memorables and all of that solves all the problems. I just don't believe current architectures will allow it

• 
  Jordan Ross (@jordan_ross_8F) reported 27 minutes ago

  I would have given this guy all my secrets and he asked me the dumbest thing. This guy I know, Ben, texted me out of the blue last week. Met him once, spoke business **** and told him if he ever needs text me. 2 years later, he texted me: "What harness are you using nowadays?" Horrible question when tapping the brain of someone you want information from to help you grow your business. If you dont know what harness is, its basically the thing that makes AI work. AI runs by having LLMs read documents and triggering workflows/solutions. Everything that goes into making those skill docs trigger to run a solution Hetzner (server) pulls information from Github (where docs live) But really...who gives AF. Im not technical. I'm a capitalist. I barely understand this ****. Im just smart enough to get it conceptually and dumb enough for it to make sense for the average agency owner who doesnt want to touch it. There are so many different questions he could have texted me out of the blue. AI is just technology. Good strategy executed like an obsessive MFer is what really counts. Ill make a boatload of $$ in this first chapter of the AI era because I am able to turn things into processes, share those processes with a 10X engineer. Set a goal ID the constraint Solve the constraint Turn that into a process that lives inside of AI. Repeat forever. The harness doesnt matter.

• 
  Harry Tormey 🇮🇪 | 🇺🇸| 🇺🇦 (@htormey) reported 28 minutes ago

  A lot of people are trying to sell enterprise AI products to VCs. Everyone from Notion to Linear, for example, seems to want to be my AI orchestration layer. Meanwhile, a lot of the actual problems we’re hitting using AI at scale are much less glamorous: GitHub falling over under load, test runners becoming bottlenecks, desktop OSes being too slow for agentic workflows, and figuring out how to run CI and E2E tests at agent scale. VCs aren’t operators, so I think there’s a growing disconnect between what sounds like an enterprise AI problem and what actually is one.

• 
  dominioscripto.eth (@septiembre_eth) reported 28 minutes ago

  3/ What just happened. OpenRouter added “Sign in with Ethereum” to its login screen. Next to Google. Next to GitHub. Millions of developers can now authenticate with their .eth wallet on the world’s largest AI platform. No password. No email. Just your Ethereum identity. (3/5)

• 
  Carlos Mora Torres (@cmora16) reported 36 minutes ago

  The core problem with Codex is not simply hitting usage limits, but that the usage meter and the product are fundamentally misaligned. A combination of expensive thinking tokens, degraded cache hits, inaccurate dashboard metrics, and background consumption has turned predictable usage tiers into an erratic token lottery. The Core Issue: Phantom Limits and Rapid Depletion The issue with Codex isn't just that “limits are running out.” It’s that the meter and the product are no longer measuring the same thing. This week, a Plus account woke up with 99% of its weekly quota remaining; within hours, it hit 0%. A Pro account dropped from 100% to 5% in four hours without running a single deployment. A 20x user burned 6% in 90 minutes without performing any heavy tasks. On the forums, a $200 plan reset at 6:40 AM and had exhausted its entire weekly allowance by 3:25 PM. Previously, hitting the cap required roughly 2.6 billion tokens; now, users are locked out at just over 400 million. This does not feel like a standard rate limit. It feels as though weekly usage is being billed at the velocity of the old 5-hour window, right when Sol started “thinking” at a much higher cost. Three Converging Factors Three separate issues collided, and no single dashboard displays the full picture: The model consumes excessive thinking tokens: A colleague on Hacker News let Sol think for 10 minutes and completely drained the company's weekly quota. That is not a refactor; that is an idle pause. Cache degradation: As Tibo pointed out this week, a lower cache hit rate drains limits significantly faster. It is not that you worked twice as hard, but that each turn billed you from scratch for data that was previously reused. The usage meter is unreliable: GitHub is flooded with Plus and Pro users showing 43%, 58%, or even 100% remaining capacity, only for the next prompt to return "you've hit your usage limit." Conversely, profiles show 0 tokens while the usage tab reads 67%. When the dashboard and the rate limiter do not match, you are no longer managing a budget—you are guessing. Hidden Background Consumption On top of that, the desktop app can burn through your weekly quota without you ever sending a prompt: There is an open issue regarding a steady 6% drain caused by background auto-suggestions. Chronicle generates summaries every 10 minutes. Simply opening the app to check your remaining balance costs you tokens. Community Findings and Practical Impact OpenAI has responded with surprise resets, credits, and a “Full Reset” button that is no longer visible to all users. Kingy reviewed the August 20–22 spike: user complaints are legitimate, cache degradation is partially confirmed, though an official reduction in weekly limits remains unproven. Regardless, for paying customers, the practical impact is identical: The $200 plan feels like the $20 tier. A 2-seat Business account hit zero weekly quota after just 19 messages. Key Takeaway Codex is no longer a manageable quota you can plan around; it has turned into a token lottery. You pay monthly for an agent that reasons autonomously, shares resource pools with Work and Excel, fails to leverage caching, and blocks access while displaying remaining quota on screen. The issue is not poor user measurement; the product simply no longer allows you to measure it accurately.

• 
  Google G-Ware Surface Technology (@GoogleWaveTech) reported 37 minutes ago

  @sama One honest distinction: each share creates a ledger commit inside StarQuest’s device-local wallet, not a new GitHub repository commit. Letting a public webpage commit directly to GitHub would require exposing credentials. Cross-device/server permanence still requires

• 
  QLI Qubic Services (@QubicLi) reported 38 minutes ago

  @RidvanSedaZtrk We'll create a GitHub issue for this.