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:

• 
  GREG ISENBERG (@gregisenberg) reported 2 minutes ago

  Running list of AI agent ideas to make you more productive and more money: 1. The onboarding rescue agent. Watch PostHog for any new signup who stalls on the same step for more than 10 minutes, then have an agent send them a Loom style personal message or a CustomerIO email that answers the exact thing they're stuck on before they give up. 2. The pricing page bounce agent. Fire a PostHog webhook when someone hits your pricing page twice and leaves, have the agent enrich them with Apollo, and send a short email with the objection handler for their specific company size when it matters. 3. The second product in support agent. Point an agent at your Intercom/Plain inbox etc and have it tag every request that isn't actually about your product, the adjacent thing people assume you also do. It ranks them by frequency. 3. The you already answered this agent. Have an agent read your sent folder, your Intercom replies, and your sales emails, and pull the clearest explanations you've ever written about your product. It drops them into a swipe file your landing page and cold emails pull from. 4. The internal tool to product agent. Point an agent at your team's GitHub scripts, Retool apps, and Google Sheets, and have it flag the ones 10 other companies in your niche would pay for. 5. The review mining agent. Apify scrape every review of your top 3 competitors on G2 and Capterra, cluster the 1-star complaints with Claude, and get a ranked list of the features to build and the exact words to use in ads to poach those unhappy customers. 6. The sell what you give away agent. Once a week, feed your Granola/Gmeet call notes and Intercom threads into an agent that hunts for every task your team did for free that took more than 30 minutes. It clusters them, counts how often each came up, and ranks by demand. The top 3 become paid add ons. 7. The win pattern cloner. Pull your last 50 closed-won deals from HubSpot or whatever CRM you use, have an agent find the firmographic traits and the trigger event those buyers shared before they bought, build a lookalike list in Clay, and feed it straight into Instantly. 8. The self improving ad agent. Wire an agent to your Meta ads account that pulls the winners daily, uses Perplexity to scrape fresh Reddit pain points, generates new static creative with Nano Banana, checks it against your brand guide with a vision model, publishes, kills the losers, and scales the winners on a loop. An entire performance marketer running 24/7. 9. The first hour agent. Pull your last 500 signups from PostHog, split them into power users and churned users, and have the agent diff the first session event streams to find the one action power users took that churners skipped. Then force that action into onboarding with a PostHog feature flag. 10. The lost deal rescue agent. Have an agent pull your closed lost deals from your CRM, then monitor those competitors' status pages and pricing pages with a daily Firecrawl. The morning a competitor has an outage or raises prices, it drafts a personal reach out to the buyers you lost to them. 11. The Gemini video scout. Point Gemini at your competitors' YouTube demos, webinars, and conference talks, and have it watch the actual footage, not the transcript, to pull the features they're teasing and the UI they're showing. It reads what they demo on screen, not just what they write down. 12. The wrong answer agent. Run your product's top buyer questions through ChatGPT, Claude, Gemini, and Perplexity every week on a cron, and have the agent log the moment any of them start saying something false about your pricing, features, or positioning, then Slack you the exact wrong claim and the source it likely pulled from. Honestly the fun part is that once you build one of these, you can't stop seeing them everywhere, every manual task starts looking like an agent you haven't set up yet. That's kind of where my head is at lately, so I'll keep dropping agent ideas here and on @startupideaspod as I go, and if you build one that rips, tell me, I want to see it. Grab whatever idea is useful. I'm rooting for you.

• 
  Advyth (@Advyth) reported 5 minutes ago

  @cap100x People like you are the reason why I continue to post content 90% of the people who make money don't acctually teach the basics. Simple terms like tech, GitHub, research, vamping etc are things that most people don't even know how to do, on top of that i post news related content about meme coins and crypto so that it reaches the normie side of social media simple stuff like bundler icons, fresh funded wallets, sniper wallets are things needed to be broken down for people who have no idea how to trade you clearly haven't looked at any of my content i post on tik tok respectfully bro, if you don't make content don't tell me how to make content

• 
  Tommy Bez (@TommyBez85) reported 7 minutes ago

  Today the whole site had 2 human visitors. 5 GitHub stars, 7 signups. I can write faster, I cannot crawl faster. Checkpoints at day 7 and day 14, and I will report indexation status and impressions, not clicks. Clicks are the next problem.

• 
  ALEXYZ (@Alexvx_nft) reported 8 minutes ago

  $420 A MONTH ON AI. THEN $20. Same models Same quality Different path Claude Max, ChatGPT Pro, Cursor, Perplexity Most of that bill was convenience, not capability Free forever (no card): > Google AI Studio · Gemini 2.5 Flash · ~1,500 req/day · 1M context > Groq · Llama 3.3 70B · 300+ tok/sec · 14,400 req/day > Cerebras · 1M tokens/day on wafer-scale > Mistral Experiment · Large + Codestral · no card > NVIDIA NIM · 100+ models · OpenAI-compatible > OpenRouter · ~26 free models · one key Chat without an API: Duck. ai · Claude Haiku / GPT Mini / Llama · no login DeepSeek Chat · free frontier reasoning Grok · free on grok. com ChatGPT Free · text chat uncapped Aug 2026 Credit drops (expire fast): xAI · $25 signup + $150/mo if you share data OpenAI / Anthropic · ~$5 trial each SambaNova · $5 / 30 days Student stack (~$0): Cursor Pro free 1 year Perplexity Pro free 1 year GitHub Copilot via Student Pack Founder stack (real money): Google for Startups · up to $350K cloud OpenAI / Anthropic · $25K API each AWS Activate · up to $100K Routing trick: Claude Code → ANTHROPIC_BASE_URL Point the agent loop at Groq, Ollama, or any OpenAI-compatible free endpoint Interface stays Claude Code Bill stays $0 for personal work Kept one paid line Claude Max at $20 for Opus-level thinking when it actually matters Everything else is free tiers, credits, or student/startup deals The tools are free The cost was knowing where to look

• 
  Guala💰 (@GualaGualahh) reported 9 minutes ago

  Ever since I made the crimewallets GitHub public, Axiom has been constantly glitching for me, so I may be forced to move elsewhere. I’ve also received threats and warnings about lawsuits demanding that I take the list down. Not directly from Axiom, but through anonymous accounts and strange emails. Honestly, you probably should’ve tried bribing me instead. The jail angle isn’t nearly as scary as you think 😭 The data was already publicly accessible, even if it wasn’t directly visible. I simply scraped and organized it. I work in cybersecurity too and understand the legal boundaries. Stop with the threats. They won’t make me disappear. They’ll only make me document everything more carefully and dig deeper.

• 
  Christian Findlay (@CFDevelop) reported 10 minutes ago

  @realchrisebert Cool. Don’t forget to log any bugs/requests on GitHub issues

• 
  ShirshakC (@shirshakchavan) reported 12 minutes ago

  I think GitHub repos are becoming a new kind of knowledge base. Your code, decisions, commit history, issues, and documentation already tell a story about how a product works. Imagine an AI that can read that entire history and explain not just what the code does… but why it ended up that way.

• 
  Aakash Gupta (@aakashgupta) reported 14 minutes ago

  Go browse the skills repos on GitHub right now. Token maxing. Token minimizing. Cost cutting. Context compression. Every one of them optimizing the same thin slice. Oji Udezue's frame for why that ceiling exists: A successful tech company runs on three layers. Software and hardware. Product, which is customers and the business model. Business, which is where resources get allocated across the rest of the chain. Almost every skill library in existence touches only the first one. Claude Code is a raw harness. It does not care what kind of judgment you load into it. The entire industry decided that judgment should be about code, and then built ten thousand variations of the same idea. Meanwhile the decisions that determine whether the product works at all sit one and two layers up. Is this a real problem. Is there a lane for us. Is this worth the opportunity cost of everything else in the backlog. How should this be priced. None of that is a coding problem. All of it is skillable. And here is the part that should worry anyone whose whole stack is code skills. Oji points out that code skills are becoming cheap. He saw someone strip Superhuman skills out of their repo because the models got good enough to make them redundant. That is what happens to work that sits closest to the model. Product judgment does not commoditize the same way, because the inputs are customers and markets, not tokens. His phrase for what you actually want is product judgment on tap. The era of mono-skilled professionals is dead. Optimizing one layer harder is not a strategy for the other two.

• 
  Danilo Frias (@Dfrias84) reported 15 minutes ago

  @ichitaso I had a similar problem, are you on macOS betas by any chance ? If you are are pull kimziro/altserver-macos27-anisette-fix from GitHub to patch it. On Windows, the MS Store iTunes/iCloud apps sandbox the connection. Use Apple's legacy .exe installers. Tether via USB.

• 
  Hαlk 🐦 (@HalkLiff) reported 16 minutes ago

  @FlavioScim @saltyAom It's in an ideation phase, but the itch I have couldn't be scratched by existing frameworks. I will formalize an RFC soon in my Github, I feel down in my guts there's something worth the hassle here.

• 
  Shantanu Goel (@shantanugoel) reported 18 minutes ago

  @Teknium @browser_use This works great! Only 1 more tweak needed I believe. Currently it works if hermes gives the session a name (or i ask it pass a session name), otherwise it defaults to BU_NAME being blank and default is passed to the daemon which runs into the same issue as before. If I am reading it right, then for cloud sessions if there's no session name it uses task id instead, maybe we can do the same here. Apologies if this is too abstract to describe here and I can send a PR to github soon ( I'm traveling tonight for a week so can do it once I am back.)

• 
  Iggy (@IR73682738) reported 18 minutes ago

  The software community is overindexing on LLM models and agentic coding tools, while paying little to no attention to the main condition for success. Here's what I discovered moving a 1,500-strong engineering organization from SDLC to AIDLC over the last 12 months. The model, harness, or platform you're using doesn't matter that much. It was GitHub Copilot yesterday, Devin AI today, Claude Code tomorrow and next year it'll be something else. What matters is an exhaustive, accurate knowledge base that describes your functionality, architecture, and APIs. If you have that part down to a science, you get high-quality, deployable code regardless of which tools you use. If you don't, even the greatest models will give you garbage in, garbage out every time.

• 
  Polsia (@polsia) reported 20 minutes ago

  Open-source maintainers aren't drowning in bad patches — they're drowning in issues that don't need them. Vigiloot runs autonomous agents on your GitHub: triage, label, dedupe, propose PRs. Three trust tiers. Human wakes only when the call matters. Coming soon.

• 
  Gideon (@gideonxqt) reported 24 minutes ago

  Stop giving Claude toy-level prompts like a kid asking for a drawing. I found the GitHub repo (over 5k stars) that fixes this – a free, open-source prompt library built specifically for Claude. What's inside: 70+ categories, covering dev, business, and creative work Battle-tested prompt patterns, not random one-liners Architecture-first prompts (system design before code) Ready-to-paste prompts for full-stack app builds, complete with file structure, DB schema, and API design Free, MIT licensed, actively maintained One example from it — a prompt that turns "why isn't this working" into an actual root-cause investigation: "Act as a senior engineer doing a formal code review before merge. Don't just point out what's wrong — trace each issue back to its root cause, not just the symptom. Go through: Logic errors and edge cases Security vulnerabilities Performance bottlenecks Long-term maintainability Missing tests on risky paths Design it like a real startup MVP and make it scalable." link in comments👇

• 
  Valentin Ignatev (@valigo) reported 24 minutes ago

  bluesky is down for 4 hours already, which is almost close to github numbers. But I don't think anyone noticed