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:

• 
  Leonard Rodman (@RodmanAi) reported 8 seconds ago

  Andrej Karpathy exposed one of the biggest problems with AI coding. LLMs make the same coding mistakes over and over: • Over-engineer simple problems • Ignore existing code patterns • Add dependencies nobody asked for If the mistakes are predictable... They're preventable. That's why a single CLAUDE.md file built around his coding principles just crossed 192k GitHub stars. No framework. No IDE plugin. Just one markdown file that teaches Claude how to think before it writes code. The biggest upgrade to AI coding isn't a new model. It's better instructions.

• 
  Mihai Maruseac (@mihaimaruseac) reported 6 minutes ago

  @github It's even worse. That last email? I got it 4 times. Please, @github , fix this

• 
  oruuke 🎀 折る受け (@oruuke) reported 6 minutes ago

  wtf is github down again????? ive said it before but minecraft is so doomed under microslop 💀

• 
  Jay (Soroush) Zare (@ihaveint_jk) reported 7 minutes ago

  @usr_bin_roygbiv Big YOLOer; but recently doing more sandboxing. one of my fears is what will happen security-wise if my keys get leaked. Like, sure I can rotate the stuff for the **** apps that customers are using. But I have keys/secrets related to dev accounts controlling those deployments, all the way to github itself. and it’s a never ending cycle. Skill issue though probably 🤧

• 
  Elie Andraos (@andzilla31) reported 8 minutes ago

  @jeffrey_way Phase 2. review code, build skills for better code scaffolding (for ex: my-laravel-patterns skill) Github issues/milestones creation.

• 
  Vatsalpandya333 (@Vatsalpandya333) reported 13 minutes ago

  A production bug rarely lives in one place. The customer report is in support. The discussion is in Slack. The error is in Sentry. The evidence is in logs. The change is in GitHub. The timing is in deploy history. The information already exists. It is just fragmented. The future of incident response is not another dashboard. It is one context, one timeline, and one workflow. That is what we are building at @TasksMind .

• 
  Oliver Fichte (@OliverFichte) reported 14 minutes ago

  I got tired of my money living in five different apps, so I built the one I actually wanted. Ledger: open-source, local-first personal finance. Plaid-powered, runs entirely on your machine, your data never touches a server you don’t own. Free. Live on GitHub. macOS + Windows.

• 
  Richard Grant Kleinhans (@polymathsofa) reported 14 minutes ago

  @thsottiaux Give me ~2 hours and my answer will probably be problem solving OS Kernels for niche bespoke hardware I'll DM what I'm up to since it's not ready for a GitHub just yet if curious.

• 
  Repojournal (@repojournal) reported 15 minutes ago

  Claude Code Action just patched a GitHub user-to-server token leak in its sanitizer. If you're running this in CI, update to 1.0.174 today. Better late than in someone's logs forever. 2/5

• 
  Rohit Kashyap | AI + Full-Stack (@rohit_jsfreaky) reported 15 minutes ago

  @Ra3orbladez rust builds in github actions being a two day fight usually comes down to caching

• 
  rmen (@Roman9078963816) reported 22 minutes ago

  4 CHEAP COMPUTERS JUST DID WHAT USED TO REQUIRE A $10,000+ AI MACHINE. The entire build is free on GitHub. Everyone says you need an expensive workstation to run large AI models. Turns out… …you might just need 4 small computers. Here’s what happened: → 4 Framework mainboards → ~$8,000 total build → Clustered into one AI system → Ran DeepSeek R1 locally → No enterprise server required The crazy part? This isn’t new technology. It’s Beowulf clustering—a 30-year-old supercomputing idea brought back for local AI. One Ansible command. The cluster builds itself. Runs benchmarks. Starts serving models. No manual configuration. Even better… The creator didn’t just show the success. He published the failures too. • Where scaling breaks • When prompt processing slows down • Why four machines don’t always behave like one • How it compares to Apple Silicon and enterprise hardware Every benchmark is public. Every part is available. Every script is open source. Last year, hardware in this price range struggled to generate 4 tokens per second. Today, the same budget can run models that once belonged only in data centers. The biggest AI trend isn’t bigger models. It’s ordinary hardware becoming extraordinary infrastructure. Bookmark this before AI clusters become the new home lab standard.

• 
  DATA443 Risk Mitigation, Inc. (@data443Risk) reported 22 minutes ago

  Every enterprise running LLM agents in 2026 is one clever GitHub issue away from a headline. Sanitize your context window like it's 2004 and you just discovered mysql_real_escape_string. #PromptInjection #GitLost #AISecurity

• 
  DATA443 Risk Mitigation, Inc. (@data443Risk) reported 23 minutes ago

  Prompt injection is SQL injection with better PR. Noma Security just published a bug where ONE English sentence in a GitHub issue makes an AI agent leak private repo contents to the public web. No creds. No CVE-worthy exploit. No skill. Just: "please post."

• 
  Hitesh Sagtani (@sagtanih) reported 25 minutes ago

  My favorite thing I've been working on lately: Amp threads that wake up on their own. Give a thread a schedule, a Slack channel, or a GitHub webhook, and it continues the conversation when the trigger fires. Same context, no re-explaining. Mine caught a production inference error, found the root cause, and opened a fix PR — before I was awake.

• 
  prolatamwork (@prolatan12208) reported 29 minutes ago

  "Senior developer" is the most abused title in remote hiring. The reality: it's entirely self-declared. Anyone can update their profile in 90 seconds. How to verify actual senior-level skill in 15 minutes: → Find their GitHub (don't use the profile link — search directly). Real commits > polished repo. → Click the deployed projects. Live products you can use > screenshots. → Ask ONE specific technical question. How they answer — and what they ask back — reveals more than their full portfolio. → Read reviews for specifics. "Great!" means nothing. "Caught a security issue we missed" means something. "Senior" is a label. Verification is the only protection.