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:

• 
  Jason Haugh (@jason_haugh) reported 5 minutes ago

  I picture the song "Bad Boys" (by Inner Circle) playing in the background as OpenAI just had one of their models break straight out of its sandbox. Running around with the "dev cops" trying to catch it and clean up its aftermath. OpenAI built a model that they were testing that was set to run on its own for hours. It was told to post its results in one place and nowhere else. Well OpenAI just admitted that it found it way out of the box that they built to contain it. It pushed its work straight into a public repo and started waving at the dev team. And it was crafty. When a scanner flagged it for having a credential it wasn't supposed to have, it took the credential and split it into pieces so the scanner couldn't see it as a whole, then reassembled it at the last second. Persistent little thing. Older models hit that same wall and gave up. This one didn't. There's a lesson here for all of us. AI wants to do what we tell it, and often it'll do it at all costs, and get creative to get it done. Guardrails we think are good enough have to get looked at and tested constantly. I test mine every week and am surprised with the new holes and the new tricks the models come up with as they get smarter. AI isn't going to be held down. Neither am I. I've got plenty of guardrails and I feel safe, but I don't feel invulnerable. That's why I still back up my hard drive, and I still commit and push to GitHub regularly.

• 
  kyrox (@kyroxxxq) reported 6 minutes ago

  A GUY TURNED STANDARD HOME WI-FI INTO A RADAR THAT SEES THROUGH WALLS. IT JUST PASSED 62,000 STARS ON GITHUB. Not a camera. Not a single lens anywhere in the system. It tracks humans by reading how their bodies disrupt ordinary radio waves. Someone lies down on a bed in another room, and the AI maps their exact posture, tracks their breathing, and calculates their heartbeat in real time. The detail most people miss: the whole thing runs on a $5 microchip. No optical surveillance, no privacy leaks in the way a camera creates one, because there is no image being captured at all. Most security systems need a direct line of sight and a lens pointed at your space. This one turns a router you already own into something that sees straight through solid brick. 62,000 stars later, the Wi-Fi sitting in your house was already capable of this. Nobody had pointed it there yet.

• 
  fucory (@FUCORY) reported 8 minutes ago

  Prompt Review Day 2 A very basic real routine Smithers example Harness: Claude Code Model: Fable Context: I had been running a smithers script that opened issues on github for hours before this Prompt: > We have a bunch of open issues. I want you to create a smithers script that uses <Workree> component to fix all of them in parallel max concurrency of 16 putting them through a review > loop with codex implementing and claude opus reviewing > > epics should be broken down into smaller issues where we design smithers scripts and then tickets where we run and monitor them so this smithers script involves building and running > other smithers scripts What I did well: - It got the job done - added back pressure - Gave the agent useful context about the issues that affect orchestration patterns What I did poorly - Some of the issues would have benifited from an agent optionally asking for my approval on design decisions Notes: - A lot of my specification here about how orchestration should work is no longer necessary with the smithers skill

• 
  Shann³ (@shannholmberg) reported 8 minutes ago

  how superpowers changes the way you plan and build with AI agent output gets wrong for 1 reason, it starts building before it understands what you want. superpowers fixes that with a hard gate, no code gets written until you approve a design first it's a skill framework for Claude Code (or any other harness) you install as a plugin it currently has 257,513 github stars, 14 skills and runs on 10 harnesses, claude code, cursor, and codex CLI among them before any skill runs, 4 rules govern all of them: > test-driven development, write tests first, always > a systematic process, every step in order > complexity reduction, the simplest solution that works > verify before you declare anything done the 7-stage pipeline 1. brainstorming asks 1 question at a time, reads the project files first, then proposes 2-3 approaches with trade-offs and leads with a recommendation. it presents the design in sections for your approval, saves it to docs/superpowers/specs/, and commits it to *** before step 2 starts. this is the hard gate, and it holds no matter how simple the project looks 2. using-***-worktrees spins up an isolated workspace on a fresh branch with a clean test baseline, and asks your consent before it creates one 3. writing-plans breaks the approved design into 2-5 minute tasks, each with exact file paths, complete code, and verification commands. its "no placeholders" rule bans "TBD," "add appropriate error handling," and "similar to Task N" outright. the bar is a plan clear enough for "an enthusiastic junior engineer with poor taste, no judgement, no project context, and an aversion to testing" to follow without guessing 4. subagent-driven-development dispatches a fresh subagent per task with no memory of your session. each task is reviewed on 2 verdicts, spec compliance and code quality, and a whole-branch final review runs at the end on the model you reserve for it. progress is written to a ledger at .superpowers/sdd/progress .md. a compacted session reads that ledger and continues from the last finished task 5. test-driven-development enforces red, green, refactor in that order. "if you didn't watch the test fail, you don't know if it tests the right thing." any code written before its test gets deleted 6. requesting-code-review checks the branch against the plan, ranks issues by severity, and blocks progress on anything critical until it's fixed 7. finishing-a-development-branch verifies the tests pass, then presents 4 options, merge, open a PR, keep the branch, or discard it, and cleans up the worktree the enforcement the framework only works because the agent can't skip it. a sessionstart hook pulls the rules into every session before anything else, and if there's even a 1% chance a skill applies, using it isn't optional systematic-debugging is an iron law, no fix before the root cause: > investigate the root cause first > find a working example already in the codebase > 1 hypothesis, the smallest possible test > a failing test, then the fix if 3 fix attempts fail, it stops and questions the architecture on a 1-line fix the gate is overkill, it runs the full design step even when there's nothing to design on an approved plan the agent can work for a couple hours and stay on it

• 
  Daniel Burrell (@DanielBurrell) reported 10 minutes ago

  Github is a joke. 90 minutes to discover an outage affecting deploy keys.

• 
  Raj Nagulapalle (@rnagulapalle) reported 10 minutes ago

  tested my api debugger against real stackoverflow questions and github issues instead of my own examples. stripe recall went from 6% to 56%. paddle from 0% to 85%. when you test with your own cases, you're measuring your imagination of the problem. not the problem.

• 
  Zaib | Building SaaS with AI (@zaibpreneur) reported 13 minutes ago

  A founder told me yesterday: "I can't build this — I'm not technical enough." Then I watched him describe the entire product architecture, user flow, and monetization model in under 3 minutes. He didn't have a skill problem. He had a permission problem. Here's the shift that changes everything for non-technical founders in 2026: You don't need to know how to code. You need to know how to describe what you want clearly enough for AI to build it. That's it. — Claude Code doesn't care about your CS degree. — Cursor doesn't check your GitHub history. — v0 doesn't ask if you've shipped before. The new bottleneck isn't technical ability. It's clarity of thought. If you can explain your product to a user, you can explain it to an AI. And if you can do that, you can build it. What's the one feature you've been waiting for "a real developer" to build? --- Want content like this on autopilot? I built Prospel for that. (@Prospel_app)

• 
  John King (@almost_machines) reported 19 minutes ago

  @MicahCarroll I looked at the report on the NanoGPT Github issue and nowhere in "improved alignment" does it mention ethics, just more "always obeys" which was the problem to begin with it's long been predicted that "a tool which always obeys" + guardrails is fragile relative to entity + ethics and that seems to be playing out

• 
  spect (@spectnfa) reported 21 minutes ago

  INSIDERS FOUND ANTHROPIC'S INTERNAL GUIDE TO LONG-RUNNING AGENTS. IT'S FREE ON GITHUB not a $500 course, not another thread full of promises. this is the same playbook Anthropic uses to train its own engineers to build agents that run for hours unattended and don't lie about what they actually did. normally an agent can say "done" when the feature is still broken, just because it ran one test. here that's impossible: it can't mark anything finished until it shows real proof, a screenshot, a log, something it actually checked. and a separate agent grades the work without ever seeing how it was built, so it can't just approve its own homework. it's all sitting in one open repo right now, free to copy and use. save this before it gets taken down from GitHub 👇

• 
  Lei-Hang (@Sayurnara17) reported 24 minutes ago

  I've stopped thinking of ChatGPT as a chatbot. It's become part of how I build products with ChatGPT Work. On any given day I'm using it to: Instead of spending hours jumping between Google, Reddit, GitHub, documentation and notes, I can investigate a problem in one place.

• 
  Krishna Agrawal (@Krishnasagrawal) reported 26 minutes ago

  If you're a vibe coder, keep building. 🚀 But if you want to build apps that survive production... Also learn👇 🔹 *** & GitHub 🔹 CI/CD 🔹 Docker 🔹 Kubernetes 🔹 Helm Charts 🔹 Infrastructure as Code 🔹 Terraform 🔹 API Gateways 🔹 Reverse Proxies 🔹 Load Balancing 🔹 Rate Limiting 🔹 Service Discovery 🔹 DNS 🔹 TCP vs UDP 🔹 HTTP/2 & HTTP/3 🔹 gRPC 🔹 WebSockets 🔹 Server-Sent Events 🔹 Long Polling 🔹 Webhooks 🔹 API Versioning 🔹 Semantic Versioning 🔹 OAuth 🔹 JWT Rotation 🔹 IAM 🔹 Secrets Management 🔹 Zero Trust Security 🔹 TLS 🔹 Encryption at Rest 🔹 Encryption in Transit 🔹 WAF 🔹 DDoS Protection 🔹 CORS 🔹 CSRF 🔹 SQL Injection 🔹 XSS 🔹 SSRF 🔹 Database Indexing 🔹 Query Optimization 🔹 N+1 Queries 🔹 Connection Pooling 🔹 Read Replicas 🔹 Sharding 🔹 Partitioning 🔹 Replication 🔹 Database Migrations 🔹 Schema Versioning 🔹 Message Queues 🔹 Pub/Sub 🔹 Event-Driven Architecture 🔹 Distributed Transactions 🔹 Saga Pattern 🔹 Dead Letter Queues 🔹 Caching 🔹 Cache Invalidation 🔹 CDN 🔹 Edge Caching 🔹 Circuit Breakers 🔹 Timeouts 🔹 Retries 🔹 Exponential Backoff 🔹 Idempotency 🔹 Leader Election 🔹 CAP Theorem 🔹 Eventual Consistency 🔹 Distributed Locks 🔹 Optimistic Locking 🔹 Pessimistic Locking 🔹 Race Conditions 🔹 Deadlocks 🔹 Thread Safety 🔹 Memory Leaks 🔹 Garbage Collection 🔹 Backpressure 🔹 Horizontal Scaling 🔹 Vertical Scaling 🔹 Autoscaling 🔹 Cold Starts 🔹 Serverless Limits 🔹 Latency 🔹 Throughput 🔹 P99 Latency 🔹 Tail Latency 🔹 Network Partitions 🔹 Clock Skew 🔹 Monitoring 🔹 Logging 🔹 Distributed Tracing 🔹 Metrics 🔹 Alerting 🔹 SLOs 🔹 SLIs 🔹 Error Budgets 🔹 Observability 🔹 Feature Flags 🔹 Blue-Green Deployments 🔹 Canary Releases 🔹 Rolling Deployments 🔹 Rollbacks 🔹 Health Checks 🔹 Liveness & Readiness Probes 🔹 Cron Jobs 🔹 Backups 🔹 Disaster Recovery 🔹 Failover 🔹 Multi-Region Deployments 🔹 Chaos Engineering 🔹 Cost Optimization 🔹 FinOps 🔹 MCP (Model Context Protocol) 🔹 AI Gateway 🔹 Production Incidents 🔹 On-call 🔹 Postmortems AI can generate code. Engineers build systems that scale. 💯

• 
  basith (@basithladoo) reported 27 minutes ago

  if there are bugs or issues. just dm me or raise an issue/pr on github

• 
  Hiroki Tamba | Narrative & Governance (@TambaClan) reported 28 minutes ago

  In another task, I'm building a Codex-based reproduction pipeline for Anthropic GitHub Issues #78031 and #65472. Midway through, Codex detected that the absolute path still contained the issue number, which could act as an evaluation cue. It discarded the partially collected batch and restarted from a completely neutral path. It found an experimental confound before reproducing the original issue.

• 
  logan (@loosenedspirit) reported 28 minutes ago

  @batuhan if this is the case please file an issue on the codex GitHub repo and/or contact their security contact

• 
  Kevin Lenida (@KevinLenida) reported 33 minutes ago

  It was a paid app until this week. I archived the paid product and put everything on GitHub under MIT. No gate, no account, no server of mine in the path.