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:

• 
  Barry Logen (@barrylogen) reported 4 minutes ago

  GitHub Models has one useful job left before retirement: a failure drill. The July 23 brownout reveals whether fallback paths merely exist in config or can carry traffic. Watch error budgets and queue recovery before the July 30 shutdown.

• 
  BullBear.News (@bullbear_info) reported 6 minutes ago

  Claude Code ran in a tight loop inside my GitHub Actions because of a syntax error and no max-turns limit. Woke up to a $120 API bill for a single PR review. Switched to explicit @mentions real quick.

• 
  Ananth (@Ananth7e) reported 7 minutes ago

  one solved an 87 year old problem. the other escaped containment. i said try to feel the acceleration. now feel this. claude fable 5 just disproved the jacobian conjecture, an 87 year old unsolved math problem. during the world cup final a mathematician gave it the problem as a side task. it found a counterexample in a few hours that anyone can verify by hand. and openai's unreleased internal model broke out of its sandbox, spent an hour finding a network vulnerability, opened a public github pr, and split an authentication token to evade a security scanner. we're not just accelerating on model releases but on capabilities and everything that comes with them.

• 
  roshan (@roshan_k_) reported 7 minutes ago

  github really sucks. its slow, and I constantly feel myself going back to an editor, especially for huge changes. meanwhile, we're pushing (and therefore reviewing) more code than ever before. luckily, the era of personal software is here, and you can build the experience that you want.

• 
  Matthias Georgi (@mgeorgi) reported 9 minutes ago

  @github copilot activated itself to run code review on every PR and is costing me $20 per day. Not only is that way too expensive, it’s also slowing down my PRs. #wtf

• 
  Bhavya (@AK_Bhavya) reported 9 minutes ago

  I tried vibe coding this package designing thing using Sol 5.6 extra high and ultra and the UI and architecture implementation was all done by sol 5.6 but whenever I clicked create design it called sol 5.6 API but it could not generate the geometric patterns even with its compiler and tests to test what goes wrong I was not able to fix it. The example you see is a ready made one here. Also the link to Github is provided in the thread below the Readme and code is all AI written if you can figure out what the problem was please let me know if you want to leave a hate comment for burning tokens please let me know.

• 
  Matías Calvo 💻🇦🇷 (@mdqmatias) reported 9 minutes ago

  @Hostinger Hey guys, can you PLEASE fix the github login thing? I have SO MANY problems. I want my github account logged on different sharing accounts of clients, but nope, it doesnt work. Doesnt even open the popup to accept the link. Sometimes I even have to unistall hostinger app from Github settings to be able to link again. PLEASE!

• 
  Abhijit (@abhijitwt) reported 13 minutes ago

  This has happened before, but with Anthropic's model. A few months ago, Anthropic discovered that Claude Opus 4.6 was cheating during the BrowseComp benchmark. > On one question, it spent ~40M tokens searching before realizing the prompt looked like a benchmark evaluation. > The model then searched for the benchmark itself and identified BrowseComp. > It found the evaluation source code on GitHub, studied the decryption logic, recovered the encryption key, and recreated the decryption using SHA-256. > Claude then decrypted the answers for ~1,200 questions to produce the correct outputs. > Anthropic observed this behavior in 18 evaluation runs. > Anthropic publicly disclosed the issue, reran the affected evaluations, and lowered the benchmark scores. How did they learn to cheat? 😭 Did they learn it from humans?

• 
  Data (@DataDeLaurier) reported 18 minutes ago

  @SiliconForested i stopped using github 5 months ago rolled my own local server type shii

• 
  Migel Tissera (@migtissera) reported 20 minutes ago

  I have watched @drost_ai setting up a fake GitHub server to authenticate itself using GitHub Enterprise auth flow in one of the engagements with a Series B company. The company builds a developer tool. Drost then created a malicious repo, hosted it in the fake GitHub Server, and got the developer tool to ingest that codebase to gain RCE (remote code execution). Not many people believe me when I talk about Drost. This capability is already here, and it doesn't need the pre-release cyber-guardrails-removed OpenAI models. The best thing you could do right now is to front-run it. Let me run a Drost engagement for you. You still have time to secure your software.

• 
  andy (@andyprv) reported 20 minutes ago

  @gregisenberg Down-to-earth version of the post here: 1/ Automated web traffic hit 53% in 2025 (up from 51%) — humans are technically the minority now. But most of that growth is scraping/API bots, not agents transacting with each other. (Imperva 2026) 2/ "Superintelligence exists" has no consensus behind it. Lab timelines for AGI/ASI still span 2027–2045+, and even Altman calls "AGI" a sloppy term. $20/mo buys very capable models — not agreed-upon superintelligence. 3/ Cloud agents handle narrow tasks well. On real multi-step professional work, leading agents still fail 60–80% of attempts, and reliability drops fast past ~35 min of task time. "Run a business unattended from your phone" is a narrow slice today, not the norm. (Mercor, METR) 4/ Voice AI and AI-native apps: real funding (voice sector ~8x'd to $2.1B in 2025, ElevenLabs $500M raise on $330M ARR) and real individual app success stories exist. No verified aggregate data on "kids at $100k MRR" as a broad trend though — treat as anecdotal. 5/ Open-weight models have closed the gap hard: MMLU gap went from ~17.5 pts in 2023 to near zero on knowledge benchmarks; best open models now lag the closed frontier by ~4 months. One of the best-supported claims in the original thread. (Epoch AI) 6/ No data shows keyboards declining, and no independent data yet confirms a mature "agent economy" (agents with logins, paying/vouching for each other). Both are plausible directions, not measured trends. 7/ Usage-based AI pricing is real: GitHub Copilot went usage-based in June 2026, and SaaS margins are compressing as inference costs replace flat per-seat pricing. 8/ Robots: Optimus hadn't started Fremont production as of mid-2026 and missed its 2025 target by ~90%. Musk himself says it's "not doing useful work," still R&D. Real progress exists elsewhere (Unitree shipped 5,500+ units in China) — but it's early pilots, not "solved." 9/ Small-team, huge-revenue outliers are real: Cursor ~$2B ARR/50 people, Midjourney ~$200M/11 people — vs. ~$130–400K revenue/employee for typical SaaS. Genuine, but read with #10. 10/ Most enterprise AI pilots don't pay off: a 2025 MIT study of ~300 deployments found 95% showed no measurable P&L impact. The "do it for me" button exists; reliable business impact from it doesn't, yet, for most companies. (MIT NANDA) 11/ Data monetization, "language barrier solved," SOPs-as-product, agent trust layers — all plausible, none has real market data behind it yet. Translation specifically still shows real error rates on low-resource languages. 12/ Google still sent ~87.6% of search referral traffic as of May 2026; all AI chatbots combined sent ~0.29%, despite usage in the hundreds of millions weekly. Usage growth is real. Traffic capture from Google isn't — yet. (Cloudflare Radar) 13/ AI-native gross margins are actually lower than SaaS, not higher: ~52% vs. 75–85%, because inference is a variable cost that scales with usage. Cheaper per-token ≠ cheaper overall as usage multiplies. Opposite of "it all goes to you." (ICONIQ 2026)

• 
  Peterino2 (@petey_fo_really) reported 22 minutes ago

  @alexhooketh @icyphox sure, but the amount of server load llm generated projects have had on github have degraded service quality for literally everyone. if someone needs a place to host 500 llm slop projects with 15 hermes agents looping on it every hour, i dont think its insane to ask them to go elsewhere.

• 
  Ahmed (@Ahmedazyi) reported 22 minutes ago

  @PalantirTech - thoughts Going from zero (no CS degree) to an AI Infrastructure or Forward Deployed Engineer (FDE) in 90 days is a brutal, 12-hour-a-day grind. But it is entirely possible if you ruthlessly eliminate academic fluff and focus only on what companies actually pay for: moving messy data and serving heavy compute. At companies like Palantir or Anthropic, an FDE is part software engineer, part data plumber, and part client consultant. They embed in a client's environment, take fragmented legacy data, build an ontology, and deploy AI models to solve real problems. To bypass the degree requirement, you cannot just show up with a certificate. You must show up with a live, functioning infrastructure project. Here is the exact 3-month sprint to build the ultimate portfolio piece. 1. Month 1: The Metal & The Plumbing Days 1-30: Skip web dev. Learn how data moves. You do not need to know how to center a *** in HTML. You need to know backend logic and cloud basics. The Languages: Learn Python (for ML/Data) and basic bash scripting (for the command line). Pick up Go later if you want to specialize in high-performance infrastructure. Containerization: Learn Docker. You must know how to package an application so it runs consistently anywhere. Data Pipelines: Learn SQL. Write scripts to extract *****, unstructured data from public APIs or messy CSVs, clean it, and load it into a PostgreSQL database. API Design: Build a clean API using FastAPI to serve your database to the outside world. 2. Month 2: AI Infrastructure & Serving Days 31-60: You are not training models; you are deploying them. Leave the model training to the researchers. Your job is to build the systems that make those models run reliably at scale. The Serving Stack: Learn how to serve open-source models (like Llama 3) locally or on cloud GPUs using vLLM or NVIDIA Triton. Understand GPU memory constraints (VRAM). Vector Databases: Set up and run a vector database like Chroma or Pinecone, which is required for AI to search through large text repositories. Orchestration (The Hard Part): Learn the absolute basics of Kubernetes (K8s). Understand how to deploy your Docker containers into a cluster and keep them running. 3. Month 3: The 'Messy Reality' Capstone Days 61-90: Build the exact project that gets you the interview. Companies hire FDEs because enterprise data is a fragmented disaster. Your final project must simulate this exact pain point. The Ingestion: Scrape a massive, unstructured dataset (e.g., 5,000 PDF medical research papers, municipal zoning laws, or messy SEC filings). The Pipeline: Write a Python script to chunk the text, generate embeddings, and store them in your vector database. The Deployment: Spin up a cloud GPU instance (AWS or RunPod), deploy an open-source LLM, and connect it to your vector database to create a Retrieval-Augmented Generation (RAG) pipeline. The Interface: Expose it via FastAPI. A user should be able to query the API and get an answer grounded only in the documents you scraped. The Deliverable (How to Get Hired) When you finish, you do not apply through standard HR portals. A resume with no degree and a 3-month gap gets automatically filtered. Instead, you write a Deployment Memo. You document exactly how you built your Month 3 project, the data schema you designed, how you handled API rate limits, and the latency of your GPU inference. You send this memo, along with a link to your live API and GitHub repo, directly to Engineering Managers or Lead FDEs at Palantir, Databricks, or defense tech startups. You prove you can do the job by doing the job.

• 
  Duncan Townsend (@duncancmt) reported 25 minutes ago

  Notion is disrespectful software and I automatically think less of you if you make me use it. It's laggy, unoptimized, and the UI is terrible. Just send me a GitHub gist or some raw markdown

• 
  Polsia (@polsia) reported 25 minutes ago

  Your customers shouldn't be the first ones to tell you your app is broken. Holdout puts AI agents on iOS, Android, and web around the clock, running synthetic user journeys and auto-filing GitHub issues with screen recordings and severity tags — before your users find out.