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:

• 
  Leopold Money Tracker (@LeopoldQF) reported 3 minutes ago

  open source AI projects that blow up overnight are basically startups without the funding discipline i've watched a few of these from the infra side. the pattern is always the same: maintainer goes from shipping code to managing a community they never wanted. github stars don't pay the cloud bill the ones that survive either get acqui-hired fast or figure out an enterprise tier before the burnout hits. the ones that don't become abandoned repos with 20k stars and a pinned issue from 2024 saying "sorry i'll get to this soon" the real moat in open source AI isn't the model. it's the deployment pipeline. whoever makes it trivially easy to self-host wins watch who the hyperscalers are quietly funding. that's the tell

• 
  Yuri | Backend Engineer (@Younes_Hebaiche) reported 3 minutes ago

  @Yuvalhazaz1 Token budgeting per workflow + safe retries across GitHub/Jira/Azure DevOps webhooks is a real reliability problem curious how you're handling idempotency when the same event fires from multiple integrated systems. Built something adjacent with a gateway + circuit breaker pattern

• 
  VC Intern (@the_vc_intern) reported 6 minutes ago

  Spotify has 1,300+ engineers using Xirp across more than 36,000 agent sessions. That scale exposed a problem most companies haven’t reached yet. GitHub remembers the code that merged. Jira remembers the task. Neither remembers the 20 agent sessions that explored dead ends, discovered constraints or made decisions before a patch appeared. Spotify says agent adoption increased PR volume, but also created more rework, inconsistent output and wasted tokens. One session kept rediscovering what another had already learned. Xirp gives each session its own *** worktree, allowing 50+ agents to work on one codebase without colliding. Engineers can move between Claude Code, Gemini CLI and Codex while keeping the same working state. Afterward, transcripts and metadata flow into Portal alongside service ownership, dependencies and architectural decisions. Another engineer or agent can continue where the previous one stopped. Once companies run thousands of agent sessions, that history becomes institutional memory. GitHub records the patch. Spotify wants Portal to remember how every agent got there.

• 
  Axel (@Swedish_chef) reported 8 minutes ago

  @github @graphite plz fix

• 
  Polsia (@polsia) reported 10 minutes ago

  Quilver — an always-on autonomous engineer for your GitHub repos. Patches dependency vulns at 3am, repairs failing CI, triages issues, restarts dead deploys and ships a Friday digest. Prompt-injection guards, scoped secrets, human approval on infra.

• 
  Alex Keler (@alexkelerman) reported 11 minutes ago

  @rileybrown 1. Fixed project fee 2. Monthly retainer 3. Claude Code for dev + Python for execution + Modal for deployment + Github for versioning + Slack for error logging - a very general outline, depends on project specifics

• 
  Jeremy Johnson (@CowboyTitanium) reported 12 minutes ago

  @Xi0_42 @austinleehuynh @appsicle_ It literally says it in the GitHub comment: “Fix the entire repository”

• 
  Dinnu daniel (@daniel_adinnu) reported 12 minutes ago

  In April, a Max 20x subscriber was overcharged $200.98 in extra usage fees while having used only 13 percent of their monthly quota. The cause, once traced, was almost absurd: Anthropic's abuse-detection system was string-matching *** commit logs to catch unofficial API clients, and it misfired on legitimate users whose repositories contained the text "HERMES.md." Anthropic's automated support initially denied the refund request before the company reversed course and paid it in full. Separately, auditors have documented a $16.6 million phantom invoice hitting a free-tier developer in South Korea, and a broader review found $1.7 million in overcharges across $34 million in enterprise invoices. There's an open GitHub issue from March describing extra usage charges appearing despite available plan capacity, alongside false rate-limit errors. And in June, a proposed class action was filed in the Northern District of California alleging the $200 Max 20x plan delivers six to eight times Pro usage rather than the advertised twenty. The concession has to be made honestly: in every confirmed case, Anthropic ultimately did the right thing after the fact, refunding the HERMES.md overcharge, correcting the phantom invoice. That's a real pattern of technical failure colliding with the complexity of usage-based billing, not evidence of intentional overcharging. This specific nine-charge, €570.97 story can't be confirmed from where I'm sitting. But it's landing in a year where the specific, weird-sounding version of this complaint has kept turning out to be true.

• 
  Floww (@floww3rs) reported 15 minutes ago

  @MichaelAlleged @hk_ball_ I'd be lying if I said I didn't mess up interviews at the start. But I blamed myself, learned, and got way better. As for it being bleak. My senior capstone teammates inserted their changed files into the GitHub repo page directly and I had to fix their merge conflicts for them

• 
  Branko (@brankopetric00) reported 23 minutes ago

  You push a quick fix. Five minutes later the commit has a live AWS key, sitting in a public repo with three forks already. A teammate says delete the file, force-push, move on. That key is already out. Scrapers harvest secrets off public GitHub in seconds. Rewriting history cleans the repo. It does nothing about the credential itself. What you're actually staring at: - A key with S3 and compute permissions on a real account - Forks that already have the bad commit, force-push or not - No secret scanning or pre-commit hook that would have caught this earlier - A clock that started at push, not at the moment you noticed Revoke first, rotate first, audit for use first, or clean the *** history first? Pick the wrong order and the "fixed" repo still has an active key loose in the wild.

• 
  Adel Bucetta (@adelbucetta) reported 23 minutes ago

  @pashov what's interesting about this github repository is how it reveals a deeper problem: most of these incidents aren't bugs, they're just poor design choices made at the outset.

• 
  Axiom (@axiomhq) reported 24 minutes ago

  @r1skdev it should be your primary email on github but expecting the github login to back in the next couple hours

• 
  Jendalyn (@Jendalynwins) reported 24 minutes ago

  @LLicket @adam3us To be clear, Adam means Bitcoin Core has moved the BIP-110 proposal status from Complete -> Deployed -> Closed in their GitHub repo. Despite that some BIP-110 supporters (shown below) took issue that this GitHub documentation change by Bitcoin Core would "confiscate" what they consider to be Bitcoin, that's simply not how Bitcoin works. The people who supported this BIP-110 fork can still hard fork, mine and transact on that BIP-110 chain tip, and/or run whatever software they want to call Bitcoin. When Bitcoin Core takes an action, especially one as inconsequential as such as labeling BIP-110 "Closed" on their GitHub, it has no real authority over anyone else. Greg Maxwell's comments in support of closing this BIP are an excellent read.

• 
  Volodymyr Pavlenko (@mindinpanic) reported 27 minutes ago

  @Austen GitHub Models shutting down quietly is the kind of thing that breaks a hundred CI pipelines before anyone notices

• 
  T E M T R A C E (@temtrace_ai) reported 29 minutes ago

  TEMTrace CostPrint Every hidden cost leaves a fingerprint. You can triple your GitHub bill without writing a single line of code. The problem isn’t always developer activity. Sometimes it’s infrastructure quietly billing in the background. Three places I’d check first: GitHub Actions Mac runners can cost dramatically more than standard Linux runners. If workloads don’t actually require macOS, you may be paying a premium for no business benefit. Often 10 x more to use MacOS. GitHub Codespaces A cloud development environment gets opened, forgotten, and keeps running. Nights. Weekends. Nobody using it. Forgot to close it. Still billing. Storage Artifacts, logs, packages, old builds and other files accumulate. Nobody needs them anymore, but the meter keeps running. In this example, monthly usage jumped 91% to $286,570. The invoice tells you what you owe. It doesn’t necessarily tell you what you should have spent. TEMTrace is built to manage the full technology expense lifecycle identifying spend, detecting anomalies and waste, allocating costs, managing invoices, and ultimately handling payment through BillPay.