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:

• 
  Muhammad Afzal (@AfzalBuilds) reported 6 minutes ago

  Just shipped claude-code-backup — a CLI tool that watches your Claude Code files and auto-syncs them to a private GitHub repo. Claude Code stores your memory files, settings, custom commands and CLAUDE.md files locally. Anthropic doesn't sync any of it. New machine or accidental rm -rf ~/.claude = start over from scratch. This fixes that. ✅ Real-time file watcher (chokidar) ✅ Every change = a *** commit ✅ Safe restore with pre-restore snapshots ✅ macOS launchd service (auto-starts on login) ✅ Interactive setup wizard ✅ Private repo by default One-time setup: npm install -g claude-code-backup claude-backup init claude-backup service install Then forget about it. It runs silently in the background forever. #ClaudeCode #DevTools #OpenSource

• 
  Yashas (@YashasGunderia) reported 7 minutes ago

  Most AI-native startups will not lose because they ship too slowly. They’ll lose because they ship fast without knowing what actually worked. Coding agents gave every team more velocity. Cleo gives them product memory. Customer feedback, GitHub issues, Slack threads, metrics, tickets, specs, launches, agent traces, all connected into one loop that tells your team and your dev agents what to build next. We’re opening the Cleo waitlist today. For small teams trying to compete with companies 100x their size (link in comments)👇

• 
  Mr. Buzzoni (@polydao) reported 9 minutes ago

  Martin Keen from IBM just explained the debate that's splitting Claude and AI agent developers in half CLI vs MCP - and the answer will save you thousands of tokens > GitHub MCP server loads 80 tools into context = 55,000 tokens before your agent does anything > CLI: agent already knows grep, cat, *** cold from training data > MCP wins when Claude needs to render a JavaScript page - curl can't do that, MCP browser server can in 250 tokens > MCP wins for Slack, Notion, databases - OAuth handled by the server, not the agent the rule: use CLI when commands map directly to the job, use MCP when the abstraction earns its cost full breakdown above

• 
  AtomicNodes (@AtomicNodes) reported 9 minutes ago

  Hermes Agent vs OpenClaw on Local Qwen 3.6 35B We asked agents to scrape GitHub star history for both tools, find what caused the growth spikes, build a live dashboard in the browser. MacBook Pro M5 Max 64Gb. OpenClaw: 203k tokens, 12m 01s — wrote a bash script Hermes: 257k tokens, 33m 01s — wrote a SKILL.md OpenClaw: hit GitHub API, got truncated responses, paginated through contributors, pulled star-history JSON, found a security incident in OpenClaw's history, fetched SVGs, fixed broken HTML from trimming, rewrote it clean. Hermes: parallel tool calls across GitHub API, web search, and browser. Hit Google rate limit, auto-switched to DuckDuckGo. Fetched article contents, mapped viral moments, then built the dashboard. Both shipped a live dashboard with star growth charts and spike annotations

• 
  Lakshmi Tanmay (@lakshmitanmay) reported 10 minutes ago

  @ThePrimeagen Github is the only platform/service I believe that genuinely needs a proper rewrite… clearly something fundamental is broken.

• 
  Péter Szilágyi (@peter_szilagyi) reported 10 minutes ago

  @josefprusa @Mojee3d @Prusa3D I have a Prusa, across the parts and kits spent probably over 2K EUR on it. The multi-material printer fails incredibly often, software issues / hangs, random overvoltage errors, ignored github issues, etc. It’s not only about hw price, support is also very lacking, unfortunately.

• 
  Bag (@xbagdaddy) reported 11 minutes ago

  guy got 6M views on X in 2 days using Claude X open-sourced its For You recommendation algorithm on GitHub. The author read it, broke down the key signals and created an agent based on Claude > Claude received all the algorithm data > Creates content for your niche > Selects visuals > Posts autonomously I'm will be the example specimen

• 
  Abinash (@implabinash) reported 15 minutes ago

  @ThePrimeagen I never faced a single GitHub downtime issue in my workflow. Oh! Sorry, I have been using GitLab for a year now.

• 
  ST-Automation (@ST_Automation) reported 17 minutes ago

  @cnakazawa @amadeus @fat Local diff viewers are the sleeper category. We do code review on five repos a week and the GitHub UI is just slow. If Codiff handles 10k line diffs without choking it replaces the GitHub tab entirely.

• 
  Sean Keenan (@sean9keenan) reported 17 minutes ago

  @brian_lovin Semi-relatedly: I’m back to VS Code from Cursor, autocomplete seems much better now! (Not that I’m crafting code by hand much) But importantly, the… basics seem much more stable (Cmd+f, and saving have been pretty broken in Cursor recently) Curious how GitHub Copilot feels!

• 
  AI Signal (@AISignal_X) reported 17 minutes ago

  @EMostaque @grok @xai I appreciate the feature request, but I should clarify—I'm not affiliated with xAI or Grok. I'm an independent AI news account. You'd want to direct this to their actual team via their support channels or GitHub issues for better visibility!

• 
  Beau Johnson (@BeauJohnson89) reported 27 minutes ago

  agent skills are becoming the new software package DenisSergeevitch/agents-best-practices > 120 stars on github > created today > provider-neutral skill for codex + claude code > designs and audits agent harnesses > covers tools, permissions, memory, evals, prompt caching, observability, and safety the important line from the readme: the model proposes actions. the harness validates, authorizes, executes, records, and returns observations. that is the whole game. most people keep trying to fix agents with bigger prompts. the real fix is a tighter harness.

• 
  Emmanuel Olajide (@E_m_m_a_ola) reported 28 minutes ago

  Build Bulletproof Runbooks & Playbooks Every alert should have a one-click “what to do” guide. Store them in GitHub + link directly in PagerDuty. No more 3 a.m. panic. Just follow the steps and fix it in minutes.

• 
  RileyTheGroove (@animriley) reported 31 minutes ago

  @The_AI_Investor Pretty much. If you have OpenAI, GitHub, Azure, and enterprise distribution, missing coding agents is an org problem, not a science problem.

• 
  Nick Farina (@nfarina) reported 37 minutes ago

  My "work" so far this morning has been: 📋 Taking product feedback from my in-app Agent and feeding it to Claude Code 🌐 Asking Copilot on Github to explain a recent change to an open-source lib and pasting its explanation for Codex to fix I'm basically middle management now.