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:

• 
  Codebender Cate™ ξ(s)=1/2s(s-1)π^(-s/2)Γ(s/2)ζ(s) (@Codebender_Cate) reported 20 seconds ago

  I need resources to find a collection of GitHub Open source arcade and casino games that can be played in the browser. I need to make sure there's no issues with copyright infringement by using the source for these games. I need true open source. Any suggestions?

• 
  Yannick Comte (@cyannick) reported 44 seconds ago

  @NimaZeighami You can already do it but it’s pain in the a**. If you look at the issues on GitHub someone was able to do it and ran beat saber. My plan is to package that into something easy to run.

• 
  Lagoon Labs (@LagoonLabsMv) reported 1 minute ago

  Pearson's anti-piracy vendor accidentally took down their own author's GitHub code repo. Paul Deitel's educational examples went dark for weeks after Link-Busters confused them with pirated textbooks. Automated takedowns hitting the wrong target again.

• 
  Hrishikesha (@HrishikeshaNFTs) reported 4 minutes ago

  @iuditg Crap. Limts worst. Drinking like juice, 20%+ for fast chat alone per day. Wasted 20+ runs failed to fix a simple github wf fix gemini pro fixed first time.

• 
  Alan (@bitforth) reported 6 minutes ago

  Harness engineering > loop engineering in 2026. Everyone is optimizing agent loops. Almost nobody is optimizing what happens when the loop fails. The biggest thing people miss about production agents is that the founder is no longer part of the runtime. During development, you’re constantly steering the agent. You retry failures, reword prompts, ignore flaky behavior, and instinctively work around bugs. Your users can’t. The moment you ship, reliability has to survive without you. That’s why the demo to production gap is really a trust gap. Production agents need to fail gracefully for a stranger, not a patient founder who already knows the workarounds. Users don’t tolerate unreliable products. They don’t open GitHub issues or file bug reports. They just… leave. If you can’t observe failures, you can’t make them reliable.

• 
  Leo Heart (@LeomathHeart) reported 8 minutes ago

  @tradematiq1 I can. You should have a desktop agent orchestrator. At least one of them (Claude code, Cursor , OpenAI Codex, Antigravity, Hermes or Goose) - I have them all. Then you should choose one specific folder on your computer or on a local server where all the agents work together. Then your agent(s) start(s) building first a documentary base (where the documentation files (usually .md) are stored) and how they should work. And you give them the task to build something (a site, a service, an automation etc.) The results are stored and create the memory for the agents and for your lab. It is good then to synchronise your scripts and documents with github or your local ***.

• 
  Raj raorane (@raorane_raj) reported 8 minutes ago

  there was this other tool before it called OpenClaw it got huge. 382k stars on github. everyone wanted it. but it had a big issue for eg: if you told your AI "check my email" and the email contained a hidden trick, it could spit out all your passwords.

• 
  Devendra Singh Mahra (@devendrasm) reported 8 minutes ago

  @_svs_ For me everything not code on GitHub issues

• 
  Chris Tidesson – e/acc (@Ravensong666) reported 12 minutes ago

  @ClaudeDevs You know what is a bit ******? I wanted Fable to fix a few security flaws in my own code - which Codex found for me because Fable even refused to find the flaws. So I gave the shitlist which GPT 5.6 wrote to Fable and Fable STILL refused. But since I only have ChatGPT Plus I couldn't let Codex do the work - the analysis alone ate up an entire 5h usage window. I mean - it's my own code, it's about OPAQUE and E2EE and similar stuff that is clearly helpful and "deredere" software work. Also it's an open source project under the AGPLv3 license and a repo I made public on github. Are you really serious? I mean - thats EXACTLY the #1 use case for such a strong model. Thats like saying "here, fork, but you can't eat spaghetti carbonara with it". Come on, you can do better, no?

• 
  Yasmine Lindsay (@yassylindsay) reported 13 minutes ago

  So disappointed in 5.6 right now! I’m a windows user… I know ******* right! The codex app has been unusable due performance issues that have shown up for me since June 12 and I’ve had a GitHub issue reporting it open for nearly as long and it’s still ******.

• 
  zyxron (@zyxr0n) reported 13 minutes ago

  4 files - 10 Claude loops - $5,500 earned. Claude code does not need another dashboard. It needs a folder that remembers what happened yesterday. The entire system is: TASK.md LOOP_INSTRUCTIONS.md PROGRESS.md outputs/ TASK.md defines the result. LOOP_INSTRUCTIONS.md defines what Claude can read, write, verify, and never touch. PROGRESS.md stores the current state, blockers, failed attempts, and next action. outputs/ holds the work so a human can inspect it. That is enough to build recurring systems for: daily project reviews, CI failure triage, PR reviews, GitHub issue summaries, documentation audits, research reports, meeting follow-ups, standup drafts, weekly retrospectives, content research. But the valuable part is not repetition. It is control. The worker produces the result. A separate verifier checks explicit pass/fail conditions. The state file records what happened. Then the system decides: stop, repeat, or escalate to a human. Without state, every run is day one. Without verification, every mistake becomes memory. Without permission limits, automation becomes blast radius. Start read-only. Then allow drafts. Then sandboxed edits. Only give it external actions after the loop survives repeated manual runs. This is not better prompting. It is a local operations layer built from four files. Prompt ↓

• 
  Kushagra Gour @css_battle (@CSSMonk) reported 14 minutes ago

  after these 10-min days quick commerce apps, amazon prime feels slow! imagine the situation where coding with AI becomes unavailable and you have to code by hand again! Even if some of us will be able to do it, we wont want to do. Just like quick commerce took away our patience coding by hand will become equally unbearable! Will AI downtimes become the next "github is down" situations?

• 
  JurixAI (@JurixAI_) reported 19 minutes ago

  We've officially registered JuriXAI Auditor as an ASP on the @XLayerOfficial AI Marketplace and we are now awaiting listing approval. The initial automated checks have already returned a PASS. JuriXAI brings automated, micro-payment-powered smart contract and GitHub repository auditing to the X Layer Mainnet. No more slow manual reviews. No more biased judging. Just fast, objective, and on-chain auditing. Here's how we are changing developer audits 👇

• 
  Nav Toor (@heynavtoor) reported 20 minutes ago

  In 2014 a Swedish engineer named Knut Sveidqvist lost a Microsoft Visio file. He went to open the diagram he had drawn a few months earlier. It was gone. Every box, every arrow, every label. All of it had to be redrawn by clicking through Visio menus again. That night his kids were watching The Little Mermaid on TV. He named his fix after the movie. Twelve years later Mermaid has 89,101 GitHub stars, 8 million users, and native rendering inside GitHub, GitLab, Notion, Obsidian, VS Code, and Confluence. Here is what the paid market still charges to draw the same boxes. Microsoft Visio Plan 2. $15 per user per month. Lucidchart Team. $10 per user per month with a three-user minimum. Miro Business. $20 per user per month. Fifty engineers on Miro Business burns $12,000 a year to draw arrows between boxes. Mermaid replaced the drag-and-drop editor with a text spec that reads like Markdown. ``` graph TD A[User] --> B[Login] B --> C{Valid?} C -->|Yes| D[Dashboard] C -->|No| E[Error] ``` Ten lines. Renders as a real diagram. Every version of Claude, ChatGPT, Gemini, and Cursor already knows how to write it. You describe your architecture in plain English and the model returns a Mermaid block. Paste it into a GitHub README. Paste it into an issue. Paste it into a pull request. GitHub renders it inline as a live SVG. No plugin. No sign-in. The paid tools shipped drag-and-drop editors. Mermaid shipped a text spec that the LLMs learned on their own. Flowcharts, sequence diagrams, class diagrams, state diagrams, entity-relationship diagrams, user journey maps, Gantt charts, pie charts, *** graphs, mindmaps, timelines, C4 architecture diagrams, treemaps. Anything you would open Visio for. Version 11.16.0 shipped two weeks ago. Because the diagram is text, it lives in your repo. Because it lives in your repo, it goes through code review. Because it goes through code review, it stops rotting. Nobody has to remember where the Lucidchart account is. Nobody has to pay $10 a month to reopen a five-year-old file. MIT license. 89,101 stars. TypeScript. The library is free forever. Mermaid Chart the company sells a hosted editor on top for teams that want one, but the core stays MIT. Somebody in Sweden lost a Visio file and refused to draw it again. Twelve years later the paid diagram tools still exist, and nobody who writes software has to use one. (Link in the comments)

• 
  Codemonger (@codemonger00) reported 22 minutes ago

  Startup Founders Pack - Claude = coding. ($20/mo) - Supabase/Convex = backend. (Free) - Vercel = deploying. (Free) - Stripe = payments. (2.9%/transaction) - GitHub = version control. (Free) - Resend = emails. (Free) - Clerk = auth. (Free) - Cloudflare = DNS. (Free) - PostHog = analytics. (Free) - Sentry = error tracking. (Free) - Upstash = Redis. (Free) - Pinecone = vector DB. (Free) Total monthly cost to run a startup: ~$20 There has never been a cheaper time to build .