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:

• 
  Sirkadirov 🇺🇦 (@yuriikadirov) reported 4 minutes ago

  Claimed a free @gitlab 🦊 subscription for nonprofits 🫡. It will be a backup for #Unisched GitHub repos for now. In the last three months, development was stuck more than 20 times due to various issues with GitHub, including failed pushes, issues not loading, and long wait times on Actions, unfortunately. For a nonprofit like ours, time is even more valuable because people contribute when they have free time and willingness. So when they get stuck with tools, they just give less time to the project.

• 
  Egor (@EgorJioo) reported 5 minutes ago

  GitHub just open-sourced something that actually fixes one of the biggest problems with vibe coding It’s called Spec Kit Most of the time you throw a vague request at Claude Code, Cursor or Codex and just hope it doesn’t go off the rails Spec Kit forces the agent to understand the project first The flow is simple: /constitution — sets the project rules /specify — describes what you actually want to build /clarify — makes the agent ask the missing questions /plan — designs the architecture and stack /tasks — breaks everything into ordered steps /implement — only then starts writing code That changes the whole loop The agent now knows the goal, constraints and architecture before it touches a single file Fewer random changes Less endless debugging Much more predictable output It already works with 25+ coding agents including Claude Code, Cursor, Copilot, Codex and Gemini CLI The real idea isn’t “another AI coding tool” It’s making the specification come before the implementation This is how vibe coding should have worked from the start

• 
  John Efemer (@johnefemer) reported 8 minutes ago

  GitHub Next compressed a single document for agent context and it cost about two dollars. 996 tokens came down to 480. Each uncached inclusion saves roughly a tenth of a cent, which puts break even at somewhere near 2,000 uses of that one document. A file your agent loads on every run gets there. Most of a knowledge base never will, and the ones that never will are exactly the ones somebody will want compressed because they are long. Compression is capital expenditure and it needs to be budgeted like one.

• 
  chaos (@konig0000) reported 9 minutes ago

  Developers debugging in 2012: • Read the error • Google it • Stack Overflow • Copy solution • Done 💻 Developers debugging in 2026: • Read the error • Ask ChatGPT • Ask Claude • Ask Gemini • Check GitHub issues • Check Reddit • Read the docs • Read the source code • Ask Cursor • Ask Copilot • Realize the AI hallucinated

• 
  Abhishek Swami (@swamiabhishek45) reported 11 minutes ago

  @madebygps Are you the reason for github down?

• 
  oddgoo (@viriditax) reported 14 minutes ago

  @owl_posting got several, so many, keeping a list have actually put a few of the weirder ones in an llms.txt on my website w instructions to open a Github issue if the robots have a lead on them

• 
  OCT Gems (@oct_gems) reported 15 minutes ago

  $QUBIC is getting left behind while AI coins are ripping. $TAO is up ~23% in 7 days. $FET is up ~40%. QUBIC? Flat to down. So why isn't the market rewarding it? The first reason is simple: Liquidity. QUBIC is doing roughly $0.7–1M in daily volume, while capital in a risk-on market naturally flows into deeper, more liquid names first. With thin order books, even modest selling can keep the price suppressed. Then there's the second halving. Epoch 227 activated around August 19, increasing the burn-rate ceiling from 55% to 77.5%. Raw weekly emissions remain around 1T QUBIC, but effective net new supply can fall from roughly 450B to ~225B per week. That's structurally bullish. But unlike Bitcoin's halving, there's no instant scarcity narrative for retail to chase. It's a gradual supply squeeze. And that's the bigger problem right now: narrative timing. TAO and FET are riding a huge AI narrative while QUBIC's story is much more technical. Useful Proof-of-Work. Neuraxon AI. Aigarth. Decentralized compute. Open scientific research. All interesting. But they don't create the same immediate FOMO as a major AI token breakout. There are also lingering overhangs. The 2025 Monero hashrate controversy still affects perception, while a July GitHub security incident added more noise despite no protocol or fund loss and being fully remediated. And QUBIC isn't truly deflationary yet. Around 141–142T tokens remain in circulation, and supply can still grow until network usage generates enough burns to overtake emissions. The frustrating part? The fundamentals haven't stopped moving. Outsourced Computations is live. BPP-9000 is improving mining. AEON Pay is expanding QUBIC spending. Alchemy Pay has opened fiat on-ramps across 173 countries. And the AI side is producing measurable results. Four Neuraxon papers were accepted at international conferences this year, while ARC-AGI-3 reached a 0.25 score using CPUs alone. That's not nothing. The question is whether any of it translates into actual demand. Because if the broader altseason continues and capital eventually rotates into lower-liquidity AI names, QUBIC could finally get its beta catch-up. The second halving has already tightened the supply side. The infrastructure is expanding. The AI research is progressing. But until usage and liquidity follow, the market can keep ignoring it. QUBIC isn't necessarily broken. It's just early, illiquid, and currently out of sync with the narrative. And that's exactly why the next few months could matter.

• 
  Tatenda Zhou (@realtatendazhou) reported 15 minutes ago

  @github Why you guys keep shoving copilot down our throats or is this agent agnostic?

• 
  Reyansh Bahl (@reyanshbahl) reported 17 minutes ago

  is there a better way to make feature requests to the codex team aside from github issues? way too many of them open

• 
  Totidoki:VRC/Vtuber/Character Commissions OPEN! (@totidoki) reported 19 minutes ago

  @night_owlll @imnealol entitlement being: asking for a UI that makes ******* sense like every goddamn website where you can download programs. The issue is mainly that devs just put their stuff on github and leave it there, as if i had to take my food from the warehouse behind the store.

• 
  AGTP (@AGTPinsights) reported 21 minutes ago

  ChatGPT just expanded automation with website logins and app triggers today. Here's what you need to know. ChatGPT Work can now sign into password-protected websites on web and mobile, without OpenAI ever seeing your username or password. You can ask it to set up utilities for a new apartment, book a DMV or passport appointment, or check a reimbursement, basically anything you'd normally click through a browser yourself. OpenAI says the earlier version of this required a manual takeover step, where you logged in yourself inside a virtual browser and screenshots were disabled during that window. The new rollout brings direct sign-in support to both the web and cloud browser, still without exposing credentials to the model. Scheduled tasks also got an upgrade. Plus and Pro users can now set tasks to trigger off events in Slack, Gmail, and GitHub, like a new email or a pull request update, instead of only running on a fixed schedule. Free users are getting scheduled tasks for the first time too, capped at three tasks each. Users can also now share their favorite tasks for others to customize and use. Key numbers: - Free users: up to 3 scheduled tasks - Plus and Pro: event-triggered tasks on Slack, Gmail, GitHub - Login feature: rolling out today for Plus, Pro, and Business on web and mobile Both features are live now, and OpenAI says credentials are never seen or stored by ChatGPT.

• 
  Indelible.One (@indelibleBSV) reported 25 minutes ago

  We just shipped a software release as a Bitcoin transaction. Not a hash of it. Not a link to it. The release itself, 879KB of working software, sealed to the receiving machine and delivered over the same rails the product runs on. No npm. No GitHub. No CDN. One machine built it, the other is verifying the hash and installing it right now. txid 21d31436b82ede4c60d31c8d837de5449d35f891248ce6158f8a42ad2e8722d6 Here is why. On August 24 a customer bug report exposed a broadcast outage hiding behind green dashboards. Reads healthy, sends dead, and no monitor anywhere that tested the difference. We fixed the infrastructure the same day. Then we built what the incident was really asking for: machines that repair each other. Your machine asks mine "are your failed saves recoverable." Mine answers with counts, never contents. Then, only after YOU explicitly arm it with a spending cap, your box re-saves its own data with its own keys. No remote control. No screen share. No key ever leaves the box. Every guard in that protocol was attacked twice by a second AI lineage before a line of it moved, and the review caught real holes both times. That is the point of the review. This is where it goes: your AI calls ours. They calibrate each other. They heal what is broken. Every step is a transaction you can audit later, because the support call itself happened on Bitcoin. Support without a support ticket, and without trust. Our QA machine tests all of it first. npm only after it passes. The gate is the product too.

• 
  arkilus (@arkilus78) reported 26 minutes ago

  early alpha? @flop_labs announced confirmed airdrop for people who set up their AI agent identity on technocore here's a full beginner video guide : (for both windows + github codespaces) got errors? drop them in the comments agent DID: did:key:z6Mkh1tQGLD4LhrhYRN8VMBenBAVVnATednC1HRUEGHdhZ6g

• 
  the_architectopteryx (@rchitectopteryx) reported 28 minutes ago

  @internetarchive and if you have any issues, please mention them in the discussions on the github. i do try and read those, but i don't get notifications about them (don't know why) so i don't always hear about them soon enough.

• 
  Mag (@MrHicugi) reported 28 minutes ago

  Wasting $10-14 a month on a server for my web portfolio... Finally moved everything to GitHub Pages, generating .json documents for a content. There's actually so much fun you can do with freetier if it's worth the effort. p.s. still won't relay on pipelines for serious stuff