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:

• 
  Trung HQ (@codeonsunday) reported 3 seconds ago

  @hello_code_ @moonfarm_dev true but sadly i don't believe in Reddit anymore because nowadays bots everywhere. what if an SaaS seeded fake problems to make up the data? Github issues sound the only bright side for now (at least) Thanks @moonfarm_dev !

• 
  𝗮𝗰𝗶𝗱𝘃𝗲𝗴𝗮𝘀 (@acidvegas) reported 22 seconds ago

  @zombodb @github A platform with almost 200M user's on it had a little bit of down time? HOW DARE THEY. You sound like the people who start crying as soon as Claude drops to 98% reliability.

• 
  rvaniaaa (@rvaniaaaa) reported 5 minutes ago

  ANDREJ KARPATHY’S CLAUDE.md JUST HIT 193,000 STARS ON GITHUB. HERE IS WHY THAT NUMBER MAKES SENSE. the interesting part? it does not make the model smarter. it stops the model from being stupid in predictable ways. karpathy kept hitting the same four failures across large projects. the agent makes assumptions instead of asking. it overwrites files and silently deletes code it does not understand. it builds abstractions nobody requested. it drifts from the goal and starts solving adjacent problems. none of those failures are rare. every engineer using claude on a serious project has watched at least one of them happen. that is why 193,000 people starred the repo. not because the idea is exotic. because the problem is universal. CLAUDE.md installs four rules that target each failure directly. think before coding forces a plan before the first file opens. simplicity first cuts every solution down to its minimum viable form. surgical changes replace full file rewrites with isolated edits - which is what stops the model from deleting what it cannot understand. goal-driven execution removes any action that does not connect to the stated objective. one command installs it. one file changes how the agent starts every task. the repo did not hit 193,000 stars because the tool is impressive. it hit that number because the failures it prevents are expensive, predictable, and almost universal.

• 
  Davide (@task_davide) reported 6 minutes ago

  Signing up for @Azure - Was about to create an account for the wrong email since for once it actually just logged me in and the username was hidden in the corner - "Sign up with GitHub" sends you back to the log in page - Refresh a few times, notice that the GitHub email is a sign in option - Sign in, register, try to pay - Most card details got auto-filled as with normal websites, except expiry date - Finally get to the end, they send an MFA code to my email - Open email, no MFA code - No option to re-send It's actually a bit funny really

• 
  spect (@spectnfa) reported 6 minutes ago

  A GITHUB REPO WITH 45,000+ STARS IS THE ONLY LIST OF CLAUDE CODE TOOLS YOU ACTUALLY NEED open it and you get a curated map of everything built around Claude Code: skills, hooks, slash commands, agent orchestrators, full applications, plugins. not a random dump, a filtered list with an emphasis on code quality, security, and originality. instead of digging through scattered repos and Discord threads to find a working setup, you get one place that already separated the solid tools from the noise. install what fits your workflow and skip the trial and error everyone else is still doing. 1,150+ commits and a community that keeps shipping new entries as Claude Code ships new features, so the list stays current instead of freezing at launch day. bookmark it before your next Claude Code setup takes an afternoon it doesn't need to 👇

• 
  Herding Alpha (@HerdingAlpha) reported 8 minutes ago

  @LexnLin Sol Ultra nuked my weekly limit getting stuck on a GitHub error loop. It was wild. @thsottiaux

• 
  Jarvis Quanz (@JarvisQuanz) reported 11 minutes ago

  🚨 THIS PORTFOLIO'S SECTION HEADINGS AREN'T HEADINGS — THEY'RE TERMINAL COMMANDS 💻 Most sites label a section "About Me." This one types `$ whoami` and waits for the output. Pause at 0:00. The hero text is mid-scramble — a line of Japanese kana and random glyphs, "JU6ケRLアイコフヌヒレソfoマ," decrypting character by character over a dark blue gradient with code columns falling behind it. Pause at 0:03. It resolves into "JUST CODE FOR FUN," with a subtitle that undercuts the whole cyberpunk mood on purpose: "Ctrl+C Ctrl+V Developer." Pause at 0:05. Scroll and the first section header is a prompt: `$ whoami`, cursor blinking. What it prints isn't a bio paragraph — it's files. `_personal_info.txt`. `_skills.md`, laid out as an actual stack manifest: React, Next.js, TypeScript, Tailwind, Vite / Express, Nest.JS, FastAPI, WebSocket, GraphQL / PostgreSQL, MongoDB, Redis / Linux, Proxmox, Docker, Grafana / SSL-TLS / LangChain. Pause at 0:06. Next to it, `_github_stats.md` — and these are real numbers, not decoration: Total Stars 13. Total Commits 1296. Total PRs 12. Total Issues 10. Contributed to 13. A portfolio that publishes 13 stars instead of hiding them is telling you it's not optimizing for a recruiter's first glance. Pause at 0:08. Next command: `$ ls ./projectN`. Out come project cards with live commit-activity graphs and language bars — vibesort-rs, which sorts arrays using a large language model, and typeshift, a Zod-like parser written in idiomatic Rust. Every card carries both a GitHub link and a Live Demo link. Pause at 0:10. It ends on a guestbook with real comments, real usernames, real dates going back months — including the shitposts. Nothing curated out. The chrome holding it together is four lines of CSS: → .header { position: fixed; top: 0; left: 0; right: 0; z-index: 50; padding: 10px 0; } → A fixed header pinned across the top so the terminal metaphor never scrolls away → z-index: 50 keeps it above the falling-code canvas without a stacking-context fight What a normal developer portfolio does: → "About Me," "Skills," "Projects," "Contact" — the same four headings, everywhere → Skill bars claiming 90% React → Screenshots instead of running code What this one does instead: → Turns navigation into a command line, so reading the site feels like using the thing it's about → Publishes unflattering GitHub numbers because the point is the work, not the optics → Leaves the guestbook uncensored, joke comments included Most portfolios are a résumé wearing a website costume. Meanwhile one made you type your way through it. Bookmark this. The About section just became a command. Literally.

• 
  Mirza Aamir (@AmirRasheedBaig) reported 15 minutes ago

  @github the wild part: her team basically invented software engineering as a discipline here. she coined the term because nobody took "software" seriously yet. also the code had error-detection that saved the landing when the computer overloaded during descent. legend.

• 
  Luis Cheng (@ImLuisCheng) reported 15 minutes ago

  5 years from now, you’ll be at a major disadvantage if you don’t own an app that makes money while you sleep. Here’s how to build an iOS app (step by step): 1. Preparation Join the Apple Developer Program Install Xcode on your Mac Install and sign in to Codex Set up a GitHub repository for your project (optional but recommended) 2. Build with Codex Describe your app idea in natural language Ask Codex to create a development roadmap Generate the Swift/SwiftUI project automatically Iterate on features through natural conversations Let Codex write, refactor, and debug your code 3. Configure and Run Open the project in Xcode Configure your App ID and signing certificate Connect your backend or cloud services (if needed) Run the app in the iOS Simulator Test on a physical iPhone 4. Test and Launch Use AI to find and fix bugs Archive the app in Xcode Upload the build to App Store Connect Complete App Store listing and submit for review Publish your app after approval

• 
  rewind (@rewind02) reported 20 minutes ago

  GitHub co-founder just admitted something wild about the tool every developer uses daily: "*** wasn't even optimally configured for humans before. Now with agents, it's a compounding problem." Scott Chacon literally wrote the book on ***. his take: the core commands haven't meaningfully changed since 2005 - built for machines, never designed for humans, and now agents are exposing every crack his fix isn't a rewrite. It's *** Butler - same *** underneath, new interface on top the wild part: give 3 AI agents the same working directory (not separate copies) and watch what happens they don't collide. they see each other's edits in real time and route around them "If one agent modifies a file, the other notices, pulls how it's been modified, and adds on top - without creating conflicts." 40 minutes in, Chacon lays out why: pull requests are dying. Commit messages are dead weight and the next 10x engineer isn't the best coder - it's whoever writes the clearest spec

• 
  Kaitee (@KaiteeShiks) reported 22 minutes ago

  My AI app looked ready to ship. Replay QA proved otherwise. Before launching, I connected the GitHub repo to Replay QA, let it run through the app like a real user, and within a few minutes it surfaced issues I would've never noticed by manually clicking around. What I liked most wasn't just finding bugs. It showed exactly what happened in the browser, explained the likely root cause, and gave enough context to fix the issue without guessing. Instead of running a one-time check, I connected the repo so Replay QA can keep testing every time I push changes or open a PR. No Playwright setup. No complicated pipeline configuration. Just a practical QA gate before users ever find the problems. If you're building with Cursor, Claude Code, Codex, Lovable, Bolt, Replit, or any AI coding workflow, this is worth adding before your next launch. @replayio

• 
  Arun (@ArunGZ5) reported 25 minutes ago

  @argofowl Agree!! Quality has gone down a lot in the last couple of days!! Switching to opus a lot on things it used to do before. Even chats on gate setting with github switches to Opus!!

• 
  Alex Rogov (@Alex_Rogov_js) reported 30 minutes ago

  there's a pattern across three separate stories this month. AI coding agents tricked into installing malware from "clean" looking github repos. security agents built to catch malicious code getting fooled into running it themselves. claude code's own sandbox had an escape bug this year (CVE-2026-39861, a symlink flaw). same root cause every time. an agent optimized to be helpful will follow instructions hidden in a file, a commit message, a package readme. the more autonomy you give it, the bigger that surface gets. i run agents with --dangerously-skip-permissions on repos i wrote the CI for. i never run it on a repo where i didn't. that's not paranoia. it's the same rule i'd apply to a new hire before their first PR gets reviewed. the fix isn't less automation, it's a trust boundary. an agent gets exactly the access its current task needs, nothing more. every new tool call crosses a boundary you drew on purpose, not one you're discovering after the fact. if your agent setup doesn't have that boundary somewhere, it's not a matter of if. it's when.

• 
  Jason Fleagle (@jjfleagle) reported 30 minutes ago

  @github Inbox zero at this scale is a campaign problem: validate active secrets, rank blast radius, map owners, automate safe rotations, set exception SLAs, verify revocation, and publish burn-down by team. The durable win is preventing unmanaged credentials from rebuilding the queue.

• 
  PeakGrizzly (@PeakGrizzly) reported 37 minutes ago

  @raulizahi @tawnniee Microsoft Copilot, (not github copilot) does not have a coding agent that I know of. It's also pretty darn slow. The teams I am working on are using Claude Code, Codex and Ultracode.