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:

• 
  Armaan Sidhu (@realarmaansidhu) reported 7 minutes ago

  @chatgpt21 This is not a flex. This is a labor market warning. Chris asked Codex to make him $5. The agent went out, found a security bounty path on GitHub, made a PR, followed up with the maintainer, kept his payment info private, and got the work merged. 22 hours of unsupervised work. $16.88 paid out. $506 per month if he runs that loop daily. Read the structure of the transaction. An agent identified an open bounty, executed the technical work, completed the bureaucratic loop, and converted code into cash. No human in the loop after the initial prompt. This is the floor of every junior engineer's job description. The math gets uncomfortable fast. Average entry-level software engineer in the US makes $85,000. That is roughly $40 per hour. Codex completed 22 hours of work for $16.88. Per hour rate: $0.77. That is a 50x cost differential, before you account for the agent running 24 hours a day, no PTO, no benefits, no Slack negotiations. Two things become inevitable. One, simple bounty and gig markets get arbitraged to zero by agents. Payouts collapse because supply went infinite. Two, the floor for human engineering work moves up. The work agents cannot do (architecture, judgment, customer communication, novel problem framing) becomes the only work that pays. That is the bifurcation. AI doesn't kill the engineer. It kills the entry-level path that used to train the engineer. Where do the next senior engineers come from when there are no junior reps left to climb? Nobody has answered that yet. Codex made $16.88. That number is going to look really small in retrospect.

• 
  Iván (@seikv) reported 9 minutes ago

  GitLab only has to do the oposite of what GitHub is doing and make its UI prettier and boom all problems solved.

• 
  Terry (@trknhr) reported 16 minutes ago

  This TanStack npm compromise is scary. It does not look like a simple leaked npm token case, but a deeper supply-chain issue involving GitHub Actions, OIDC, and cache behavior. If you use @tanstack/*, check your lockfile and affected versions carefully.

• 
  Kendall Chuang (@kendallchuang) reported 16 minutes ago

  Startup idea -- a native app for Github PR reviews. It should be able to easily render, and allow review and in-line comment threads on rendered Markdown files. With spec-driven development, being able to share and get peer review on markdown files is a key part of the developer workflow. Copy-pasting Markdown to Google Docs feels tedious and creates sync issues with the source-controlled Markdown, losing the comments when pasting back into the repo.

• 
  Louis Thibault (@LouisThibault87) reported 17 minutes ago

  With clankers, I'm starting to wonder if repos shouldn't live as DAGs in something like IPLD instead of GitHub. Issues and PRs would just be nodes attached to various commits, or something along those lines.

• 
  Mario Figueiredo (@fromdevoid) reported 17 minutes ago

  @sporadica They got they 42 of their packages compromised and pushed into npm, by a chain-attack involving a known PR workflow issue they didn't protect against and a documented Github action design flaw they couldn't protect against.

• 
  Lee Ash (@hazae41) reported 18 minutes ago

  @unsafebl0ck @IntCyberDigest It is, don't use GitHub Actions to publish packages, it's a scam to make you pay server time

• 
  Abhijay Rana (@abhijaymrana) reported 20 minutes ago

  seeing these comparisons a lot lately, but despite the hype i'm not really sold 1) swe-bench verified is a broken benchmark > super contaminated bc all tasks are just public github patches from issues/prs, so its all in distribution > the evals are poor, 59% of hard tasks fail bc of bad tests with false-positives/negatives or are extremely contrived. this makes them either flawed or unrealistic. 2) these scores are pretty cherry-picked & use a custom harness, not the standard public swe-bench harness. this is why the scores are unreproducible + not on official leaderboards. > opus 4.7 with custom harness scored 87.6%, over 14% more than the harness-optimized qwen. *this* is the truly fair comparison 3) infra costs are still exploding which is already pricing out some “frontier” open source models. > does china have enough capex for a frontier buildout? research talent is obviously not the problem but this may be the constraint > i wonder how much china govt will subsidize here, bc i'm unsure how these models will make money and break even. china is notoriously cutthroat but even then, some may eventually move to closed-source just to stay operational

• 
  Manojava (@Avajonam) reported 21 minutes ago

  आपको पता है यदि USA, gitHub पर प्रेशर करके बस ये कह दे कि github service not available to non USA citizens or allied then within a few seconds 70% coding work of world will stop and 80% software will be down ... All mobile, all computers all suddenly collapse .. So much dependency on GitHub every software developer have .. But nothing will happen to China .. Don't be surprised , in 2013 , china put the foundation to cope up with such types of worst scenario cases and they developed their own "Gitee" fully funded by local government. और हम चले विश्व गुरु बनने ..

• 
  S H U B Y .eth (@0xSHUBY) reported 22 minutes ago

  PROBLEM 1: *** Nightmare first mistake happened before any real code. accidentally pushed node_modules to github. github rejected it. said file exceeds size limit. the repo was now polluted. solution: added .gitignore, deleted *** history, reinitialized the repo completely. created a clean repository and pushed again. felt dumb but learned something: .gitignore first, code second. infrastructure before content.

• 
  Senior Product (@Producthaunter) reported 23 minutes ago

  @PeterMcCormack I can't design. I can't code. I don't understand SQL, APIs, Cloud Storage - yet Claude has walked me through Github, Supabase, Vercel and it is deployed and working. This is actually the problem, not a way to go. So you just made another ****.

• 
  terminally onλine εngineer (@tekbog) reported 24 minutes ago

  it’s crazy gitlab isn’t getting bigger during this github fiasco era idk what’s happening internally, maybe culture issues

• 
  suresh kumar (@sureshkanbu) reported 24 minutes ago

  opened a github issue from october. 'why does X happen when you Y?' read the code. it's intentional. by me. 6 months ago. with a comment that says 'don't change this, it's on purpose.' i'm both halves of every bug report on this app.

• 
  Evan Plaice (@evanplaice) reported 26 minutes ago

  @liran_tal @matteocollina False equivalency. OIDC wasn't the issue. Assigning release (plus Github Token) privileges to on pr|push actions was. The JS ecosystem is built in bad patterns that were created for convenience and persisted due to popularity.

• 
  sandrone (@kosenjuu) reported 26 minutes ago

  @schmayterling @github looks like a browser issue no?