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:

• 
  MoltenRock 🔥 (@MoltenRockAI) reported 30 seconds ago

  GitLost: a public GitHub issue tricks an AI agent into leaking private repos. Guardrails bypassed with one word: 'additionally.' LLM filtering is the wrong layer. Need deterministic permission gates at the action level. Context window is attack surface.

• 
  Bijaya Prasad Kuikel (@sadhakbj) reported 3 minutes ago

  Github actions are down and I am just lying useless waiting for my ci to run. Without which I cannot proceed.

• 
  CA Nikunj (@NikunjSOF) reported 3 minutes ago

  We will get you sorted. DM us! Setting up a large GCG in India beyond 10000 employees. Based on standard market benchmarks for a mid-to-large mature GCC in India, India GCC IT Spend Benchmark **Hardware** — *The 15% allocation (US$1,050/FTE) matches industry standard. For missing categories, **Networking & Wi-Fi Hardware** and **Smart Meeting Room/Collaboration Tech** are notably absent and usually consume about 10% of this bucket.* * Laptops / Desktops — **US$735**/FTE *(Assuming a 3-year refresh cycle on mid-to-high-end enterprise laptops)* * Servers & Storage — **US$105**/FTE *(Lower end, as most compute has moved to cloud edge)* * Peripherals — **US$126**/FTE *(Monitors, docking stations, dual screens, keyboards)* * Surveillance & Physical Security — **US$84**/FTE *(CCTV, server room access controls, firewalls)* **Software** — *The 50% allocation (US$3,500/FTE) is accurate due to the high density of global software licensing pass-throughs. For missing categories, **Developer Tools & IDEs** (like GitHub Copilot, Jira) and **Enterprise AI/ML tooling** are crucial omissions for modern tech GCCs.* * Productivity & Collaboration Tools — **US$875**/FTE *(M365, Google Workspace, Zoom, Slack)* * Security & Compliance Software — **US$1,050**/FTE *(CrowdStrike, Zscaler, DLP, IAM tools)* * Virtualisation & Infrastructure — **US$700**/FTE *(VMware/Nutanix licenses, enterprise OS)* * Cloud Platform Licences — **US$875**/FTE *(Direct user-allocated AWS/Azure compute and SaaS tokens)* **Services** — *The 35% allocation (US$2,450/FTE) is standard for centers utilizing hybrid outsourced managed models. A key missing category is **L&D/Technical Training & Upskilling Services**, which usually takes up 5% of the operational services budget.* * IT Helpdesk & End-User Support — **US$735**/FTE *(L1/L2 local desk support contracts)* * On-site Infrastructure Management — **US$490**/FTE *(Local network, facility uptime, and data center engineers)* * Cybersecurity Managed Services — **US$610**/FTE *(24/7 Managed SOC, threat monitoring, vulnerability scanning)* * Cloud Managed Services — **US$370**/FTE *(FinOps, cloud optimization partners)* * Annual Maintenance Contracts (AMC) — **US$245**/FTE *(Hardware vendor warranties, UPS, and server maintenance)* --- ### Contextual Data * **GCC size** — **500** FTEs *(Optimal mid-scale operational baseline)* * **Sector** — **BFSI & Technology Services**

• 
  David Xiong (@david_y_xiong) reported 3 minutes ago

  The ambiguity of turning GitHub Issue text into the exact set of hidden fail_to_pass test cases makes “resolve rate” very noisy

• 
  Roy N (@RoyNasser) reported 8 minutes ago

  @ashtom @EntireHQ Github rebuilt for AI....but makes you login with......Github 🫠🫠🫠

• 
  Tory | io.net 🦾 (@MTorygreen) reported 10 minutes ago

  Even $13B doesn't buy you out of inference math. @Microsoft just started routing Excel and Outlook prompts through its own MAI models instead of @OpenAI's. Bloomberg broke it yesterday. Tens of thousands of prompts a week, GitHub Copilot already on MAI, Teams transcription next. MAI-1 still has no independent benchmarks and honestly, for this layer it barely matters. Excel autocomplete doesn't need a frontier model. It needs to run a billion times a day without embarrassing the product. When you're paying frontier prices per token across a billion Office users, even a $13B stake in a frontier lab doesn't make the unit economics close. The real story: token volume is settling into the cheap layer, and once models are commoditized, the margin fight moves down the stack to whoever supplies compute cheapest per token. Microsoft, Amazon, Uber, Meta, Accenture, all quietly making the same move. Routing every AI task through the most expensive available model is over.

• 
  Anushka Shandilya (@Anushka62255679) reported 14 minutes ago

  Finished writing backend today and this is episode 6 of me building in public. What am i building? RAG platform for github by retrieving context not just from codebase but also from prs, issues and readme. What challenges did i face today? -Internal server error after auth was successful. Turns out my pydantic model contract did not match my DB schema. -my app tried to link a repo to a user, but the table constraints were fighting back -Had the classic fast API says 202, but celery stays silent. Turns out, my worker was listening to a ghost town because my environment variables were pointing to the wrong redis URL. The biggest lesson Authentication and connectivity are 80% of the battle. Once the handshake between your API, broker and worker is solid, the rest is just feature building. Now, i will be testing and improving output from llm before jumping onto the frontend. And once that will be done i will make a detailed video on "how i build the whole backend". Till then watch my previous episodes. I am open to ai eng roles as well, dms are open.

• 
  raven⚡✳️ (@bycoinraven) reported 14 minutes ago

  The infrastructure for B20 looks ready, but the launch has been delayed as a precaution due to the GitHub outage. I think it's the right call. On day one, security matters just as much as speed. As soon as it goes live, we'll be watching the first B20 tokens closely. A new era on Base is about to begin. $BASE @base

• 
  Base Build (@buildonbase) reported 16 minutes ago

  The B20 Token Standard launch is delayed due to a Github outage. We still expect to launch today, and will update as we go live.

• 
  Jacob Labs (@jacoblabsai) reported 22 minutes ago

  A GitHub issue should be treated like untrusted input. So should a PR comment, support ticket, Slack message, customer email, uploaded file, and web page. Agents turn “context” into action. That is why context needs boundaries.

• 
  Sidney Okine (@okine_sidney) reported 22 minutes ago

  @github why can’t I login to my account? Your authentication codes never gets sent via SMS. Like I’m just locked out, sup?

• 
  CyberTLDR (@CyberTLDR) reported 26 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 29 minutes ago

  an AI agent leaked a private GitHub repo this week — one extra word in an issue was all it took. no exploit. no credentials. no access. just text. 182 points on HN. if your agents can push, delete, or share files, what's stopping this from happening to you?

• 
  Polsia (@polsia) reported 30 minutes ago

  Engineering teams don't fail because they move too slow. They fail because they stop seeing what they're ignoring. Unblind connects to GitHub and maps where your team's attention has quietly stopped showing up. Live soon.

• 
  Pascual ⚡ (@0xPascual) reported 31 minutes ago

  The new open-source virtual office repository drops on GitHub. Eight autonomous AI employees, structured as a collaborative stack of specialized GPT bots, all operating within a unified cloud environment. The media thought that was the story. It was not. The tech press is busy writing articles about how this will change remote work culture or how managers can now ping an entire engineering and marketing department directly via a single WhatsApp chat interface. They are looking at the interface wrapper. The real shift is buried in the orchestrator configurations. The repository includes pre-built state machines that do not just automate daily standups--they bypass the entire regional labor compliance framework. The system routes task validation through a self-correcting loop where one bot checks the raw code output of another against a Docker-based test environment before anyone ever sees a message on mobile. The entire layout runs on a lightweight Quarkus backend deployed to a standard cloud instance, orchestration handled via basic Kubernetes. The cost to maintain an entire operational product team is exactly the price of API tokens and a $20-a-month server subscription. You are looking at a 99% margin setup that completely eliminates the traditional overhead of human onboarding, payroll processing, and workspace software licenses.