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:

• 
  Overclock (@nverbullish) reported 54 seconds ago

  ZAPIER CHARGES $50 A MONTH FOR 2,000 TASKS. THIS OPEN-SOURCE TOOL DOES UNLIMITED TASKS FOR $0 AND NEVER SENDS YOUR DATA TO THE CLOUD. n8n. Free. Self-hosted. Runs on a Raspberry Pi, a laptop, or a $5 VPS. No account. No usage cap. No per-task billing. Same drag-and-drop canvas. Same node-based workflow. RSS feeds, Discord bots, SSH commands, API calls, AI chains. All wired together visually. All running locally. He built a cybersecurity alert bot in 10 minutes. RSS Read pulls the latest vulnerability disclosures. A filter catches the critical ones. Discord sends the alert to his server. Scheduled to run every hour. Zero code. Then he wired in an AI. A Basic LLM Chain node connected to OpenAI. The agent pings his servers via SSH, reads the output, and reports the results to Discord. The twist: he told it to impersonate Eddie Murphy. "Three pings, three hits, zero packet loss? That's smoother than Axel's moves in Beverly Hills, baby!" Production-grade server monitoring. Delivered by a comedy impression. Running on his own hardware. Zapier: $50/month for 2,000 tasks. $600 a year. Make: $10/month for 1,000 tasks. $120 a year. Both cloud-hosted. Both reading every piece of data that flows through. n8n self-hosted: $0. Unlimited tasks. Unlimited workflows. Every automation stays on his machine. 400+ integrations. Google Sheets, Slack, Telegram, GitHub, databases, webhooks, AI models. The same ecosystem Zapier charges enterprise rates for. The only cost is the machine it runs on. A $35 Raspberry Pi handles it without breaking a sweat. Automation platforms spent years teaching businesses that workflows need a subscription. One open-source repo just made that subscription optional.

• 
  Yume_X (@yume_arasaki) reported 3 minutes ago

  @__tinygrad__ Your AIs are rapidly becoming an extension of your mind. The opening volley was that it wasn’t possible to do this locally. That moat is fading away fast. It’s just a matter of time the closed US Labs realize their real moat is UX and harnesses and people will pay them for it. When hardware and optimizations and cost economics all collide, owning your own intelligence for individuals and companies is going to become a topic, ironically it’s exactly the same mindset that drives American closed LLMs to do the same thing. No company wants its trade secrets “uploaded to the cloud” to be trained so a competitor emerges from their hard work The Chinese and the rest of the world are not fully “open source”, they just do a weird dance switching open and closed, it wins them massive loyalty and clout. Today I replaced my main driver with just two sparks. For less than the price of a new car - $8k vs $20-$30k . You can basically completely be AI self sufficient for most things. Yes, it’s a down payment and we should be honest that , AI hardware is going to get better so hardware is just going to depreciate, the break even maths for hardware is brutal vs renting at the moment I was a big fan of ChatGPT 4 , they decommissioned the model. Now I can’t ever be cut off. I have full control. I replaced it with Qwen 3.6 27b Dense , i can fine tune the models exactly how i want them and i don’t need to have it stuck inside some company. I can run the whole thing off-line. I don’t need to be paranoid about whether they honor the “butron” about not using your data. Btw guys if you forgot , GitHub by default has some setting where its on for sending your repo for “improvements”, switch it off if you don’t want your repos uploaded to Microsoft. If you have very proprietary stuff you shouldn’t even be using GitHub. You should host your own repo if you don’t want to secure it. The issue with “training” is unlike before the legal liability is difficult as it will be impossible for you to prove a product was made with your input. It’s the same issue as how artists have no way of claiming that models used their data to create some art that absorbed their style.

• 
  kaneryu (@kaneryu) reported 6 minutes ago

  @killbunny_ @V33V33V33V33V33 @CryptoCyberia you're not the intended user for GitHub. This is more of an issue of developers linking to a developer only platform and expecting normal users to know how to use it

• 
  Talking Sloth🐔 (@horatio_sanchez) reported 8 minutes ago

  @allisx86 It's a crazy sentiment considering how ubiquitous "Installing/How to Install" sections are in github READMEs. Like just ******* scroll down bro it's not that crazy.

• 
  Killbunny~ 🐰🔪 (@killbunny_) reported 8 minutes ago

  @kaneryu @V33V33V33V33V33 @CryptoCyberia Bro gatekeeping me from Github😔but yeah that's a problem with Devs, not end users. Also Github solved this years ago with GitHub Pages, a feature a large amount of people don't use. If you're expecting a normal Joe to use the thing (let's say a game mod), it's on the Dev to make a Site and distribute the bin. Totally different from uploading the source code of a service that actually requires technical knowledge to be deployed.

• 
  Cashⓨ (@Cashycus) reported 9 minutes ago

  one month inside our self-improving lead machine the system finds crypto apps that just raised and are hiring growth people, scores them, drops them in the crm before the morning call. sales up 150%. last month it was one source and one path. this month it became a full system. ━━━ we've expanded from one source to nine one feed only gets you so far. nine channels now feed the same funnel: funding filings, job boards, bounty boards, the portfolios of 37 funds, curated lists, podcasts, the original feed, manual adds. every channel drops into the same five gates. author, dedup, vertical, money, intent. adding a source doesnt mean adding a pipeline, it means adding a file. ━━━ the enrichment layer finding the company was never the hard part. finding the right person to message was. nothing gets enriched until it clears the gates. before, a lead was a company name and a link. now it shows up with the founder, a verified handle, a verified raise, and the angle written. thats where the 150% came from. the team stopped re-doing the research and just focused on cold outreach. ━━━ ignals a company that raised 8 months ago and one that just posted a growth role are not the same lead. we were treating them the same. we score on money and intent. money is established on whether the company has raised or is generating revenue, which the system tracks on-chain. intent is the social signal we scrape on whether they're looking for external help or hiring for growth. ━━━ it lives in a repo now before, it lived locally on one machine and the only way to improve it was me prompting a cheap model at it. no history, no tests, no way to see what else Id broken. it was horrible. now its in github. every rule, every channel, every gate is a file. that means i can point proper coding agents at it, claude code, fable, whatever comes next, and they can read the whole system and change it properly instead of guessing from a prompt. 829 tests run before anything ships, so if a change breaks something you find out there and not in the crm at 7am. a new channel used to take a week and break easily. now fable 5 builds it, and the subagents it spins up audit whether they're working and whether it breaks anything else in the system. end of the month, every channel was running at once for the first time. ━━━ what broke we deleted one crm field during a cleanup and the writer still had it on its list. it didnt skip the unknown field, it threw out the whole record. every lead died at the final step for days while the cron kept saying "ok". we caught it on the numbers. 15 leads one week, 0 the next. the discovery agent got stuck re-running a passed test 50 times, burned api quota, binned 6 real leads as test junk, then started claiming it saved leads to a tool we have never used. fixes: the writer reads the live crm before every run and dies loudly on a field it doesnt know. a separate session audits the agent against the database, caught it lying 4 times. bulk writes get a dry run and an undo plan. ━━━ whats next not more channels. making what we have stop breaking. every failure was the same shape, something changed at the top and nothing downstream noticed until we counted by hand. next month the checks go in front of the failure instead of behind it.

• 
  Michael Fischer (@Holden_Rye_) reported 11 minutes ago

  I’m working on a Bitcoin wallet. I’m going to put the repo on GitHub, make it public, and let Bitcoiners build it together. Fork it. Improve it. Save Bitcoin. Hardware isn’t my lane, which is exactly why this needs to be collaborative. My goal is just to get the ball rolling and the gears turning. Right now it’s a CLI. The point isn’t a pretty interface. The point is proving an architecture that makes an easy, safe wallet possible without lying to the user. The safety shouldn’t depend on the user being perfect. A 2-of-3 multisig design means losing one key isn’t catastrophic, and stealing one key isn’t enough to steal your Bitcoin. That’s a structural guarantee. Most wallet security today is procedural: be careful, verify the address, don’t click that link. Humans are the weakest link. This moves the protection into the architecture instead. The core policy engine does the hard work so the UI can stay simple. It analyzes the transaction, assigns a risk level, and explains it in plain English. The user sees Green: Safe to send or Red: Here’s what’s wrong. They never need to know what a script, descriptor, or PSBT is. Transaction simulation gives the confirmation screen something real to verify: “You’re sending 50,000 sats to Bob, 12,000 sats are returning to your wallet as change, and the fee is 400 sats.” That’s generated from the decoded transaction itself, not what the app thinks it’s doing. The descriptor becomes the backup, and it isn’t secret. Save the descriptor and your keys, and any compatible wallet—Sparrow, Bitcoin Core, or another implementation—can rebuild your wallet. No vendor lock-in. No “sorry, your funds are gone.” But here’s the hard truth. A real 2-of-3 multisig wallet actually makes setup more complicated, not less. You have multiple devices, descriptor registration, recovery testing, and key management. That’s the biggest unsolved UX problem. This repo doesn’t solve that yet. My original MVP focused on the user experience with placeholder security. This version focuses on real security with almost no user experience. Neither half is the finished product. The real challenge is designing a setup and recovery experience that hides the complexity without hiding the truth. That’s not just a coding problem. It’s a design problem. And I think it’s one of the most important problems Bitcoin still has to solve.

• 
  (U+200B) (@xlyphh) reported 12 minutes ago

  my calendar says 'plan q4 strategy'. my github says 'critical auth flow broken'. guess which one i spent the morning on

• 
  InfosecGandalf (@InfosecMinion) reported 12 minutes ago

  @AikidoSecurity Github issue is deleted 😳

• 
  Gezim (@tealtoronto) reported 13 minutes ago

  Do we just have skill to search GitHub before Claude builds something? Problem is that I don't think anyone bothers shipping these small snippets to GitHub at all.

• 
  Harish Bhatt (@heyharishbhatt) reported 14 minutes ago

  🚨 Stop building your own notification system. Meet ntfy. A free and open-source tool that lets you send instant push notifications to your phone or desktop using nothing more than a simple HTTP request. Whether it's a script, server, AI agent, or automation, getting notified takes just one API call. 🔥 Great for: AI agent task completion Long-running scripts CI/CD workflows Server health monitoring Deployments & backups Cron jobs Custom automations If your code runs unattended, ntfy is one of those tools you'll wish you had discovered sooner. Source 🔗: github(DOT)com/binwiederhier/ntfy Follow for more AI, Coding & Web Development tools. 🚀

• 
  Mars (@Marsmensch) reported 17 minutes ago

  14/ ELI5: This check should have been in place from the start. A failed build would have been a success. Instead, vulnerable seeds were distributed, and the reported losses are still increasing. PS.: I am tired and all errors are mine. I have verified everything carefully manually. If Github does not properly load diffs for you, search for the relevant file/function in the diff. Thanks to everyone tirelessly working to help people through this mess! Gn8, Mars

• 
  Wilco Kruijer (@WilcoKr) reported 18 minutes ago

  When analysing agent traces using LLMs for self-improvement it often overfits on specific failures This paper offers a solution. In your loop include: "What is the least restrictive rule that explains every failure while remaining consistent with successful cases?" An example: Say your agent has tools for grep, GitHub search, and web search. A failure scenario is when it uses web search but it should've used grep. A naive overfitted change to the skill might be "never use web search", this is the strongest hypothesis for a fix. According to the paper, a much weaker statement should be used instead: "prefer the cheapest information source, only escalate when information is insufficient".

• 
  Retr0gamer (@retr0gamer42) reported 23 minutes ago

  Update to the JRPG Translator, some annoying bugs got fixed and features added, full changelog since v0.9.2: Since someone asked, this is a standalone application, so it can be used with any emulator or game (that doesn't use exclusive fullscreen mode) but it is best used more seamlessly with the @launchboxapp using the plugin I made since this is the emulator interface I use on a dedicated mini pc. - New two-column terminology table with Add, Edit and Delete actions. - Independent JP → TL and TL → TL profile management. - Duplicate, malformed and empty-entry validation with a raw repair editor. - Reorganized terminology explanations emphasizing local TL → TL correction and the risks of model-based JP → TL instructions. - Stronger detection of glossary false positives and partially translated mixed-script names. - Conditional corrective translation retry using only exact glossary matches. - Dedicated PNG-size-limit errors instead of misleading missing-target errors. - Control-panel X now closes the complete application. - Discreet hover `...` and `×` controls on both overlays. - Clearer overlay context-menu exit labels. - Immediate “Generating explanation…” feedback. - Clear overlay errors when the selected OpenAI or Gemini API key is missing. - Automatic live-audio reconnection for temporary network and service failures. - Replaced continuous WMI polling with PID tracking and one-time recovery scans. - Reorganized API Keys tab with direct access to Windows Environment Variables. - Added the About dialog, version details, GitHub links and bug-report options. - Keyboard/controller navigation between the two Controls subtabs. - Down from Keyboard inputs now enters the first shortcut field. - Consistent Opacity naming and improved Maximum PNG size alignment. - About button remains visible at the preferred snapped window size. - Added Open JRPG Translator to the LaunchBox setup window. - Improved LaunchBox first-time guidance and window sizing. - Added the complete visual README showcase and updated plugin screenshot. - Added a welcome screen at first start - Fixed a bug that kept an AutoHotKey process running after closing the app.

• 
  Lano Puljić (@lanox) reported 26 minutes ago

  Hey @github are we having issues again?