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:

• 
  Pulseon (@pulseon_dev) reported 4 minutes ago

  Agent benchmarks are the new LLM benchmarks. Every lab publishes SWE-bench scores. The leaderboard turns over every few weeks. What it measures: 2021 GitHub issues. Production codebases don't look like that.

• 
  Rakhul (@rakhulkarthick) reported 5 minutes ago

  @testingcatalog the swe bench pro claim needs checking... their github shows they tested on a 25 issue subset not the full 2294 benchmark if you run the full suite the score drops hard. saw this with qwen too when they cherry picked docker-only cases

• 
  Matt Slotnick (@matt_slotnick) reported 5 minutes ago

  @preetkailon if you can’t get developers to use github copilot and you can’t get business users to use copilot cowork, how is smashing them together supposed to solve the problem?

• 
  mert (@im_mert_) reported 6 minutes ago

  @BuildWithOm Resume is theoretical, while GitHub is practical. Doesnt matter if you say you have expert knowledge in, lets say C++, but havent had a commit since 2 months ago (which's an issue). And it also does not matter if you consider yourself a C beginner, but wrote your own allocator.

• 
  Fundl (@fundl_us) reported 11 minutes ago

  @paulg one reason we built fundl around verified stripe/github data instead of self-reported metrics: a single lie would crater the whole platform's credibility. the cost of one broken promise is asymmetrically high when trust is your core product.

• 
  Luka (@mladluka) reported 12 minutes ago

  35+ agents, 24h+ autoresearch loop optimizing an NLP imbalanced class problem with 1.000.000+ LOC PR 10 research agents scraping the internet: arXiv, GitHub, Kaggle, Medium, etc.. and saving findings to research.md. 10 implementation agents adapting research to the concrete domain problem, training models and running evals, logging to logs.md. 10 feedback agents performing full error analysis cycles and proposing next architecture iterations to feedback.md. So far improved the existing production real-time SOTA model by 5 points

• 
  Rahul Chaudhary (@ghost_ray2050) reported 12 minutes ago

  @emergentlabs Hi @emergentlabs , have sent the mail to support team 2 days back about the issue related to save with GitHub. I haven't received a single acknowledgement from team saying that it's been looked upon. No update on the same. My work is stopped, not getting response from the team.

• 
  Averrouz (@TheRiceFather) reported 16 minutes ago

  @atmoio I ask my self sometimes what is worse: 1. EV range anxiety 2. Social media fatigue 3. Token limits anxiety 4. Github being down anxiety 5. Code working but it does not make sense rage I usually get 2-3 at a time thank good not all 5 imaging road raging in my EV because i was debugging code.

• 
  Ahmed Alnagar (@ahmed0alnagar) reported 22 minutes ago

  Cancelled my GitHub Copilot subscription. ~300 credits per prompt. 300 × 5 prompts/day × 4.5 days = 6,750 credits. That's nearly an entire 7,000-credit ($39) plan gone in less than a week. Add frequent errors and unreliable outputs, and it's hard to justify. @Copilot @github

• 
  Francis Marion (@InternetBureau1) reported 23 minutes ago

  @AlongKumar1 @RTHztk7 @PooWorldOrderr Indian products are typically spaghetti coded from Github. Especially if they're h1b visa workers. Companies usually have to rehire the Americans they fired to fix what a bunch of indians with fake degrees did.

• 
  Nitin Rachabathuni (@nrachabathuni) reported 28 minutes ago

  GitHub's down *again*. It's not just reliability issues; it's a symptom of wider infrastructural decay in big tech, and we've normalized it. How much does this dependency shape our definition of "developer" now? ---END--- #software #github

• 
  Peak (@Mujibv5) reported 28 minutes ago

  I spent a week researching what developers actually hate about modern dev tools. Not opinions. Data. I analyzed 500+ Reddit threads, GitHub issues, and HN comments across Neon, Supabase, LangChain, and PostHog. Then I ranked every pain point by impact (frequency × hours wasted). The results are wild 🧵

• 
  U.S.A.I. 🇺🇸 (@researchUSAI) reported 28 minutes ago

  🇺🇸 The First Order Consequence: GitHub Copilot experienced service degradation as requests relying on affected GPT-5.2, GPT-5.3-Codex, GPT-5.4, and GPT-5.5 models were impacted by an upstream provider issue tied to the Responses API, reducing reliability for individual developers’ coding assistance during the affected window 🇺🇸 The Second Order Consequence: Developers and teams using Copilot encountered slower or less consistent completion behavior, prompting workflow adjustments such as pausing automation-dependent tasks, switching temporarily to alternative tooling, or retrying prompts, which in turn reduced short-term productivity and group momentum until service stability returned 🇺🇸 Discernment: The degradation highlighted operational dependency on upstream API health for model-backed features, reinforcing the value of monitoring, redundancy in development workflows, and incident-aware planning based on prior patterns of upstream service variability 🇺🇸 Reasoning: As routing depended on the degraded Responses API path, affected model requests were more likely to fail or underperform, which shaped real-time behavior by encouraging retries, fallback strategies, and more cautious reliance on live assistance while service quality was impaired 🇺🇸 Judgement: The incident indicates a controlled but meaningful period of growth decay driven by external dependency, with recovery likely restoring baseline collaboration capacity once upstream stability returned, while increasing organizational discipline around resilience and contingency planning

• 
  Lyrie.ai (@lyrie_ai) reported 30 minutes ago

  Three zero-days in Windows are sitting unpatched right now. Microsoft's response wasn't to fix them. It was to delete the researcher's account, suspend their GitHub, and threaten criminal charges. Nothing makes researchers go quiet faster than watching that play out.

• 
  Pacote (@pacote__) reported 30 minutes ago

  @thsottiaux Using Codex on WSL, if Codex is installed in a non C: drive, it just crashes when you open it, its been like this since launch, was never fixed and the github issue has been open for sooooo long.