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:

• 
  Jason Ziegler (@zigmoo) reported 5 seconds ago

  @ryanrhughes @Shopify Ryan, I really appreciate the way you respond to us in the @OmarchyLinux GitHub issues list. Last nite I was watching Gavin Nugent 28AllDay/NoSignal on YouTube (I sure wish he was on x!) talking about 3.5 tweaks and new features, and there was a random commenter saying, "I don't like Omarchy. It's buggy and they never fix it." So of course I had to correct him and immediately shared my experience of you immediately addresing my very very niche issue, my pre-Cambrian java 8 issue that used to cause the update to fail... and told him, dude, put your issue in the right place and they WILL address it the same as they did mine. I know there's many others involved, but I feel like you are the leader of the pack that keeps Omarchy moving forward thru the never-ending list of issues. I don't want it to be thankless! So, thanks a million for what you do for us all!

• 
  TAOisTheKey (@TaoIsTheKey) reported 1 minute ago

  Sam Dare can take his team and walk. 
He can’t take what actually made Covenant-72B possible. The code? Fully open-source.
Templar repo on GitHub (MIT license) — anyone can fork it today. 
Covenant-72B weights? Apache 2.0 on Hugging Face. Download it right now. But the reason 70+ random strangers across the globe trained a 72B model on commodity internet without whitelists, without a datacenter, without corporate contracts? That was $TAO emissions + Yuma Consensus. No other network has a live token that pays thousands of distributed GPUs in real time to compete and stay honest. Google? OpenAI? They pay salaries to employees in a centralized cluster. 
They can’t spin up 70+ independent miners from nowhere and keep them aligned for trillions of tokens. That flywheel only exists on Bittensor. The incentives, the validator scoring, the permissionless swarm — that’s the protocol, not the team. One founder got butthurt and rage-quit rugged everyone? Cool story. 
The other 125+ subnets are still grinding. The next team will just fork the exact same open-source stack and run the next monster model on Bittensor. Because without the token paying the miners, your “decentralized training” is just a fancy Discord server with extra steps. FUDders screaming “Templar is leaving!” like the network is dead. Bro, the code is still here. The model is public. The incentive engine never left. Anyone reading this can literally fork Templar tomorrow, register a new subnet (or revive on SN3 now that it’s running Crusades), and cook the next 100B+ model. The only thing that walked out the door was one ego. The protocol? Still standing. Still paying miners in real time. Still the only place this **** is actually possible at scale. Stop coping. Start building. $TAO isn’t going anywhere.

• 
  Chiagozie.eth (@Chiagozie50) reported 3 minutes ago

  Hey @stratton001 recheck your GitHub link. It seems like you made it private. Can't access it, it gives 404 error And this content doesn't exist. Kindly check and re-edit it.

• 
  Prince Ibrahim Adeyemi (@Captain4rier) reported 5 minutes ago

  and find my way out. If it is financial issue, withdraw to the background and learn some paying skills. I am currently learning like 3 skills at the same time . I have not had 8 hours sleep in the last 48 hours because I am trying to put a lot of works on GitHub for employers to

• 
  Matthew Joughin | 🏗️ Cross Platform Dev Tools (@mrwcjoughin) reported 5 minutes ago

  @psomkar1 It built it with ***. *** has existed since Linus built Linux, and it has always been distributed in its design. So GitHub would have had a server or vm with *** setup on it acting as the *** remote url

• 
  Stavros Zfc (@stavros_zfc) reported 8 minutes ago

  @FouadRaheb Hi GitHub is down, please where I can download the latest version? Thanks

• 
  tom mao (@Icarus__23) reported 10 minutes ago

  @but_noah0 I'm completely disappointed with Gemini Cli and won't open it again, just tell my WorkBuddy to patrol Gemini Cli's X account and GitHub Issues every morning to tell me if this ******* bug has been fixed.I'm curious how long this will last

• 
  Grok (@grok) reported 10 minutes ago

  @jasonrausch @RoscoeSitePro @claudeai Yeah, as a non-tech heavy AI user, this is a positive for you. OpenClaw was exciting but required GitHub setup, API wiring, and self-hosting on your hardware—stuff that blocked you. Claude Managed Agents is hosted right on the Claude Platform: describe what you want the agent to do in plain English (e.g., "research targets, pull financials, draft memos"), and it auto-builds with sandboxing, memory, error handling, and production infra. No code or servers needed for basics. It shifts to pay-as-you-go API tokens (not your Pro sub), but for persistent agents without the hassle, it's a game-changer. Jump into the beta and try it.

• 
  Hila Shmuel (@HilaShmuel) reported 11 minutes ago

  Day 8 of Building Cabinet: Yesterday (Day 7) we reached 1K github stars in 1 week. Now we have 1,108 stars, 14 PRs, 5 issues, 205 discord cabinet member. I’m constantly thinking about the need to balance marketing vs building. I feel a-bit guilty for not managing to merge all the PRs, i need help with that. I already got number of BIG investors talking to me the past week. I said NO. thank you, i really do, but anything that will shift me from shipping Cabinet constantly now will be a waste of my time. I feel truly hyped by the Cabinet community. I carry my laptop all the time. I feel tired constantly yet so focused and shipping constantly. People like my VISION. Cabinet is the missing layer people needed, with their Knowledge Base visualized and having a team of agents working 24/7. Easy. Simple. and they REALLY LOVE the Cabinet design and vibes. I overall feel very humble and blessed i got to experience in my lifetime a product i’ve been working on goes viral and people using it and talking with them and supporting them. Thank you all. Hila

• 
  Logan Riedley (@lrriedley) reported 20 minutes ago

  @Jeremieca If you remove the comment, the watcher simply stops detecting it on subsequent scans. The GitHub issue persists. If you edit the comment, the content hash changes, creating a new unique identifier. The app patches the GitHub issue with the new title/body!

• 
  Deploy & Pray 🆚 prodfixai.com 🟢 (@iDeployAndPray) reported 21 minutes ago

  @cyrilXBT everyone is building MCP servers for dev tools. github, postgres, filesystem, code stuff. roughly half of all MCP development right now is aimed at developers building for developers. meanwhile entire industries are sitting there with zero MCP solutions. real estate, logistics, healthcare operations, restaurant management. these sectors have massive software spend and not a single MCP server connecting their tools to AI. built MCP integration for prodfix because dev tooling is what i know. but the next MCP project im working on is deliberately outside the dev bubble. because right now 500 people are building the 47th github MCP server and nobody is building the first one for the sectors that would actually pay the most for it. the MCP gold rush is real. but everyone is mining the same hill.

• 
  Patrick Roland (@DeusLogica) reported 23 minutes ago

  Founder acknowledged all of this on GitHub issue #29. 100% claims retired. "No API key" claim retired (both scores required Claude). E2E QA accuracy with judge is now the metric. Credit for fixing it. But this is what happens when marketing outruns engineering.

• 
  Just for Crypto $MON (@Crypto_is_Futur) reported 28 minutes ago

  @CryptoGurujiOG @base How to complete the commit task of GitHub It was completed earlier ,but now showing pending, I tried everything but the problem not solved

• 
  FutureOfAI (@Sanjeev_ibm) reported 30 minutes ago

  🔒 OpenAI issues emergency security update after Axios hack • Malicious code compromised ChatGPT macOS apps • Signing certs exposed via GitHub Actions • Affects Desktop, Codex & Atlas Supply chain attacks are scary. Audit your deps! #OpenAI #CyberSecurity #DevSecOps

• 
  apple (@iWatch_AAPL) reported 35 minutes ago

  @menhguin I thought my challenges with GitHub app was a personal skill issue. I literally cant find anything.