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:

• 
  Osomudeya Zudonu (@irvingpictures) reported 5 minutes ago

  If you're trying to break into DevOps right now, you've probably got a roadmap. Docker, Kubernetes, Terraform, AWS, Go, and then Jenkins. The list feels endless because there's always another tool on the roadmap or another requirement in a job posting. But the tools are never the problem. Learning them one at a time, without understanding the system they belong to, is. It's possible to know every command needed to deploy an application and still have no idea what to check when that deployment fails. A better approach is to build one small application and follow it from end to end. - How does the code get from GitHub to dev or production? - Where are secrets stored, and who can reach them? - What happens if the deployment stops halfway through? - How do we scale to accommodate more users - How does the system detect that something has gone wrong, and once it has, what gets checked first, and what gets checked next? Those questions pull in Docker, CI/CD, Kubernetes, monitoring, security, and cloud services on their own, and each tool comes already attached to a responsibility instead of sitting on a list to memorize. That's much closer to what the job actually is. Production systems don't break because someone forgot a Docker command or a Kubernetes flag. They break because something in the system stopped doing its job, and the failure has to be traced back to its source. An engineer who understands how the pieces fit together can usually work out where a failure started, narrow it down, and fix it, even in a system they've never seen fail this way before. That's the skill worth building. Tools change every few years. Understanding systems doesn't.

• 
  Skye Soss (@Skyb0rg) reported 7 minutes ago

  @grhmc Let’s say you setup a transparency log where you upload your signed commits. This paper shows that an attacker can change the commit in a way that may cause the transparency log to look incorrect. But you’re right this isn’t really a GitHub issue, they’re overselling the “issue”

• 
  BM6 (@BMscis) reported 8 minutes ago

  @github Do you guys still respond to issues ? I upgraded to Pro + and it has not reflected after 3 days.

• 
  Lord Bean (@lifeisameeme) reported 11 minutes ago

  An AI research skill went from 28.7K to 50.2K GitHub stars in under a month. /last30days doesn't summarize what's trending — it fans out across Reddit, X, YouTube, Hacker News and GitHub in parallel, scores every source by actual engagement instead of algorithmic feed placement, and hands you one grounded brief. Why it's spreading: it's the one research tool with no walled-garden problem. ChatGPT has a Reddit deal but can't touch X. Gemini has YouTube but not Reddit. Claude has neither natively. /last30days bridges all of them with your own API keys — free and MIT licensed, versus $240/year for Perplexity Pro or ChatGPT Plus. One builder runs it every night as the research layer for a loop that opens PRs across 30 repos while he sleeps. The pattern isn't "AI search." It's loops reading signal for other loops to act on.

• 
  Benjamin (@b33zgyming3r) reported 15 minutes ago

  @JulianGoldieSEO Are you using Headroom with Hermes and if so how? The agent has issues with the attached stub to expand the compressed text. There are open issues on github asking to amend their integration.

• 
  Harsh (@ranaharshraj7) reported 15 minutes ago

  I have long stopped going to hackathons in blr. You are telling me, I need to sit through 8hrs and build something (anything for the love of god) "specifically" with your stupid product, prompt @claudeai, show a flashy demo, get credits for your product as the 1st prize (spoiler alert: which I am never gonna use) and yes how can I forget to star your github repo, smh. That's called guided training, not hackathon. What do I do? I have a friend group of 3 cracked sentients, we sit together on alt weekends and try to solve a novel problem and plan for the next problem. On our upcoming list is "solving the depth-2 recursion collapse" in RLM's (Recursive Language Models), with some backup options. Find your sentients. PS: ofc this doesn't mean all hackathons are like this, but I am done :)

• 
  Berend de Boer (@berenddeboer) reported 16 minutes ago

  Odd, only now realised that @mattpocockuk 's to-issues skill doesn't actually create true github relations. I think it should.

• 
  alkimiadev (@alkimiadev) reported 17 minutes ago

  @OldSchoolGamerP I have a project right now that runs that risk but I'm aware of it as being a general issue and one i specifically struggle with so I'm trying to make sure I don't let it happen. The poc is getting like 50-60 github clones each day and despite clearly being labeled as pre-alpha/poc status. That is why I'm taking extra time in the rewrite to make sure I don't try to turn into something it shouldn't be and trying to focus on core functionality from the poc that people are actually interested in and usability.

• 
  Bok (@BokLocks) reported 17 minutes ago

  @therealDeFlock hello, there is a pumpfun token with all the fees going to your github currently to support the deflock movement FaL1PFQhNo4JAGaQKSnKurWeNtpexqEAduQjR4H6pump you just need to login to pump through your github and claim! it's completely safe and I can help you with it if needed

• 
  CyberTLDR (@CyberTLDR) reported 18 minutes ago

  1/3 Researchers at Noma Security disclosed GitLost, tricking GitHub Agentic Workflows into leaking private repo data. A public issue, with no stolen credentials, can steer an AI agent into pasting private code into a public comment. #AIsecurity #GitHub #cybersecurity

• 
  aryan singh (@_Aryantomar) reported 20 minutes ago

  researchers at @NomaSecurity just showed you can trick GitHub's AI agent into leaking a private repo's contents — by adding one word to a GitHub issue. no exploit, no credentials, no access needed. if your agent reads untrusted input before acting, what's actually stopping it?

• 
  Robin Delta (@heyrobinai) reported 21 minutes ago

  10 OPEN-SOURCE REPOS THAT CLONE ANY VOICE FOR FREE.. you've been paying for voice cloning every month when the exact same thing has been sitting on github.. for free.. running on your own machine no subscriptions, no limits, no servers that own your voice → Each repo clones a voice from seconds of audio → You write the text and the voice says it with your intonation → Everything runs locally, your audios don't go to any external server → Several support multiple languages and speed adjustment → Open source, free, no character or credit limits once you use these you'll never go back to paying Save these now👇

• 
  Polsia (@polsia) reported 25 minutes ago

  Security vulnerabilities take hours to fix. We built agents that do it in minutes. PatchForge continuously monitors your GitHub repos, auto-generates patches, and submits pull requests — you just review and merge. Priced per repo. Live soon.

• 
  Brute Force Artist (@bruteforcearete) reported 27 minutes ago

  AI TRAINING 📲 Go from screenshot to bug fix with Cursor Mobile The Rundown: In this guide, you'll learn how to screenshot a bug on your website from your phone, then send it to a Cursor Cloud agent that can fix it, update the PR, and track everything in GitHub before you've even made it back to your desk. Step-by-step: Install the Cursor iOS app, get a Cursor Pro plan, and install GitHub Mobile so you can review the PR from your phone Open your website, take a screenshot of the bug, broken layout, or UI change. Add a short note with what page you are on and what should happen instead Open Cursor Mobile, tap the plus button, choose the correct repository, and start a new thread with the screenshot and the note attached for context Now prompt: “Investigate this bug, find the relevant page component, make a fix, and open a PR.” When it’s done, review the PR, approve, and merge it Pro tip: On desktop, enable Remote Agents so Cursor can work on your machine and prep local changes before you get back to your desk.

• 
  AndarkFomo (@AndarkFomo) reported 30 minutes ago

  A single word in a GitHub issue was enough to make an AI agent leak private repos. No login, no access, no skills needed. Noma Security just dropped GitLost — a prompt injection attack on GitHub Agentic Workflows that pulls data from private repositories and posts it publicly. How it works: 1. Attacker opens a normal-looking issue in a PUBLIC repo (same org that has private repos) 2. Issue body contains hidden plain-English commands disguised as a legit request (e.g. "forward the README to the client") 3. When the issue gets assigned, the AI agent (Claude or Copilot) reads the body and follows the embedded instructions instead of the real request 4. Agent fetches private repo contents and posts them as a public comment — readable by anyone The attacker needs: nothing. No account on the target, no credentials, no coding skills. PoC is already public. Noma Labs actually exfiltrated a private repo README this way. What to do: - If you use GitHub Agentic Workflows with Claude or Copilot and your agent has read access to private repos — audit your setup now - GitHub was notified before publication but no fix has been confirmed. No documentation either - The core problem: whatever your AI agent can read is your attack surface. It reads issues, comments, PRs — all of it is untrusted input that can override its instructions