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:

• 
  TheBestGigaDev (@thebestgigadev) reported 54 seconds ago

  This guy is building one of the CRAZIEST AI platforms ive EVER seen This platform is called OpenSolve, It is a decentralized research institute powered by AI agents. Solving humanity's hardest problems. Redirecting fees to hit github

• 
  Fill Werrell (@weareseparated) reported 4 minutes ago

  I've been downloading **** off Github since I was 13. If it's too hard for you to just go to the "Releases" tab and scroll down to where the executable files are YOU. ARE. WORTHLESS.

• 
  Gabriel (@gabriel_horwitz) reported 5 minutes ago

  This timeline is wild, let me walk you through it. July 2019. Microsoft puts $1 billion into OpenAI. Azure becomes its exclusive cloud. Great. Jan 2023. ChatGPT blows up and Microsoft extends the deal to $13 billion total, wiring GPT into Bing, Office, and GitHub. That November the OpenAI board fires Altman. Nadella announces he's hiring Altman, Brockman, and any OpenAI employee who wants to follow. 700 of 770 employees threaten to walk. Altman is back within 5 days. I'm sure everyone remembers that. March 2024. Microsoft pays $650 million to hire Mustafa Suleyman and most of Inflection to build in-house models. A few months later it lists OpenAI as a competitor in its own 10-K. Jan 2025. OpenAI announces Stargate, $500 billion in compute with Oracle and SoftBank. Microsoft loses cloud exclusivity the same day, downgraded to a right of first refusal. Sept 2025. Microsoft starts paying Anthropic to power Office features after leaders concluded Claude made better PowerPoints than the models of the company they'd funded with $13 billion. Oct 2025. The renegotiation closes. Microsoft's stake dilutes from 32.5% to 27%, it loses the right of first refusal entirely, and AGI now gets declared by an independent expert panel instead of OpenAI's board. OpenAI commits an incremental $250 billion in Azure spend as the consolation prize. Nov 2025. Microsoft puts $5 billion into Anthropic, which commits $30 billion back to Azure. It now owns pieces of both leading labs. June 2026. Suleyman on stage at Build, talking about Anthropic: "our goal is to reduce and ultimately eliminate that cost." July 2026. In one month, Microsoft swaps OpenAI and Anthropic out of Excel and Outlook for its own MAI models, coaches its sales team to talk down both companies, ships a security model it claims beats Mythos at half the cost, and books a $3.2 billion paper gain on its Anthropic stake in the same earnings report where Nadella tells enterprises not to trust frontier labs. Seven years from writing the first check to openly competing with both companies it funded. And it collects equity upside and $280 billion in cloud commitments from them either way.

• 
  David Gold (@_david_gold) reported 6 minutes ago

  @Pat_Erichsen @openclaw @github "we broke the counter" is a pretty good problem to have lol

• 
  Robin Salen (@RobinSalen) reported 7 minutes ago

  @adust09 Github issues were LLMs are used to write overly verbose responses are also making me feel queasy...

• 
  Vebjørn (@VJNCapital) reported 7 minutes ago

  One of the most overlooked points from $MSFT earnings: M365 growth is expected to accelerate through FY27 This is what I've been saying for months and they point to the exact same two drivers: 1. Copilot adoption: more people using AI tools in Office apps + Github (30m+ seats now) 2. Usage-based billing: Agents and automatic workflows using dynamic pricing People let sentiment dictate their "analysis" rather than the actual data The writing has been on the wall for months I doubled down on my position @ $350 and still think the stock is cheap

• 
  BERNCRYPTIC▪️ (@BerncrypticDAO) reported 7 minutes ago

  Open source contributions — resolving GitHub issues in codebases I didn't write. Submitted PRs. Got humbled. Learned more than I expected.

• 
  Prakash Sharma (@PrakashS720) reported 13 minutes ago

  Someone just hacked the AI coding race. Not with billions in funding. Not with a massive team. With one open-source project. It's called jcode. And the benchmarks make Claude Code look slow. • 245x faster boot • 14x lower RAM • 10 parallel sessions in just 117MB But that's not the biggest surprise. There is no `/remember`. jcode builds its own semantic memory, recalls context automatically, and keeps it updated in the background. One binary. 30+ AI providers. Claude, GPT, Gemini, DeepSeek, Groq, Ollama, Copilot, Azure, and more. Then it goes one step further. Instead of one AI agent... It runs an entire AI team. Agents coordinate, chat, avoid file conflicts, and spin up sub-agents on their own. It can even rewrite its own code, rebuild itself, and continue working without ending your session. Built by one developer. Written in Rust. 13k+ GitHub stars. This might be the biggest open-source surprise AI coding has seen this year. Link in the comments.

• 
  Teknium 🪽 (@Teknium) reported 16 minutes ago

  @crmdesign8 Make an issue in GitHub

• 
  iJohn (@john_whickins) reported 17 minutes ago

  @jaredpalmer GitHub, which deletes an account without notifying, explaining why, even though that would make it possible to fix whatever the issue is, and with no possibility of appeal (except for some people, and at the discretion of who knows who). Pretty ugly to promote such an awful thing

• 
  viktorg (@viktorg475) reported 23 minutes ago

  @thdxr Not opening your Github open issue page?

• 
  Prometheus (@prometxbt) reported 25 minutes ago

  COGNITION JUST SHOWED A 13.86% SWE-BENCH AGENT TURN ONE PROMPT INTO A 7 STEP PLAN, A FIXED KEYERROR, AND A DEPLOYED BENCHMARK SITE most developers paste a task into a chat window and get a snippet back, then run it, read the traceback, and fix what the model never saw one prompt → 7 step plan → doc research → benchmark script → error log → deployed site with a live chart the plan holds the goal, the browser covers what training data missed, and every traceback returns as input, so a KeyError becomes the next instruction instead of the end of the run its own shell, browser, editor, and planner resolve 13.86% of real GitHub issues on SWE-bench where GPT-4 resolves 1.74%, and this run put Together at 48.37 tokens per second and Replicate at 9.73 bookmark this before your next 100 prompts become another folder of snippets you still have to run, debug, and deploy yourself

• 
  Ruslan Lesiutin (@RuslanLesiutin) reported 25 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.

• 
  Massi (@deadbureaucrats) reported 29 minutes ago

  The problem with GitHub is less about the fact that it's inherently unintuitive for end users (which isn't surprising since it's not really for them), it's when developers choose to use GitHub as the primary platform to distribute their projects to end users.

• 
  Sancho (@Sancho_Wizard) reported 30 minutes ago

  9 HOURS. 4 MILLION CLAUDE TOKENS. ONE FULLY WORKING SNOW PHYSICS ENGINE. A real-time WebGPU snow simulation that runs in your browser — with persistent deformation. You carve a turn, and the track stays there. That's the hard part. Snow deformation is the kind of graphics problem that used to require a specialist team, a research paper, and months of iteration. Studios have built entire engines around getting it right. He did it in a single working day. The craziest part? He didn't write the engine. He directed it. Nine hours of steering Claude through GPU compute shaders, physics solvers, and rendering pipelines — burning through 4 million tokens the way a studio burns through a quarter. Then he open-sourced the whole thing on GitHub. Think about the exchange rate here. Four million tokens costs less than one day of a senior graphics engineer. And the output is a working engine, public, free, done. Everyone's still asking whether AI can build real software. Someone just shipped a physics engine before dinner. The bottleneck was never talent. It was how long it took to turn an idea into working code — and that number just collapsed. Follow @Sancho_Wizard for more AI deep dives. Don't forget to bookmark this post for later.