GitHub's decline: 8 outages, AI-generated code, and the future of storage

exactly this. agent scale is a whole new thing. this is like the early days of the web, when we started crushing back ends with self-service transactions. agent business is the new ebusiness, and our backends are going to need to be rebuilt and rearchitected to accomodate that change. the next 100x is coming.

mapping end to end processes and thinking about what your new pressure points will be what we need to get good at, and quick! tools like value stream mapping can be super helpful here.

The backend load explosion with agents reminds me of what happened to banks with the shift to mobile banking. A technology unlocked a behavioral shift (dopamine powered), which increased demand on services by 10-40x factor. My talk track back then was something like this: "Once upon a time, you got into your horse and buggy and brought your paycheck to the local bank to deposit, maybe twice a month. At some point you got direct deposit and the bank built a website, so you checked it there instead. Then you got a phone and for some reason banks are seeing people check their accounts on payday 40 times to see if the check has cleared." True story. And as James said, it required a massive, multi-year rebuild/re-architecture to accomodate. That was just dopamine-fueled curiosity and a device that let us do something on the train or standing in line because we got bored after 7 seconds. This is going to be so much bigger.

Great post from Anurag. It's like whacamole. AI agents unlock productivity... which bursts through the damn and floods the valley below (github, render, and soon more!).

GitHub has quietly suspended new individual subscriptions for Pro, Pro+, and Student plans around April 19 2026. On the surface, it looks like a simple capacity decision. In reality, it exposes a deeper shift in how modern developer tools are being used and stressed. The root cause is not just growth. It is behavior. Agentic workflows, where developers rely on AI to iteratively generate, test, and refine code, are consuming far more compute than traditional usage patterns ever did. What used to be a few API calls is now continuous interaction with high end models, often running in loops. That kind of demand compounds fast. This pressure is not isolated to GitHub. It ties directly into broader infrastructure constraints, especially on Microsoft Azure, where a large portion of this compute is provisioned. When capacity tightens at that level, product decisions upstream start to change quickly. New users cannot subscribe to higher tier plans for now. Free trials were already paused due to abuse. Usage limits are becoming stricter, with token based throttling and session caps becoming more visible to end users. We are entering a phase where demand for AI assisted development is outpacing the supply of compute needed to support it. That gap forces trade offs. Limits, pricing shifts, and access controls are not edge cases anymore. They are becoming the norm. If you rely heavily on these tools, it is time to think a step ahead. Assume constraints will tighten, not loosen, and plan your workflows accordingly. #GitHub #AI #DeveloperTools #SoftwareEngineering #CloudComputing #Azure #AIDevelopment #DevOps #TechNews #GenerativeAI #LLMs #BuildInPublic #Productivity #Coding #FutureOfWork #ScalableSystems #Engineering

GitHub was designed for humans. AI agents are breaking it. I ran a batch of 40 AI coding agents against a single GitHub repo last week. Within 90 seconds: rate-limited, merge conflicts on every branch, and three token revocations. The architecture assumes a human opens a PR, waits, reviews, merges. Agents don't wait. Cloudflare just shipped a Git platform built for this exact problem. 𝗧𝗛𝗘 𝗕𝗢𝗧𝗧𝗟𝗘𝗡𝗘𝗖𝗞: GitHub's API rate limits and merge queue assume sequential human workflows — agents operate in parallel at machine speed 𝗧𝗛𝗘 𝗦𝗛𝗜𝗙𝗧: Cloudflare's platform treats concurrent writes, branch isolation, and agent-scoped auth as first-class primitives, not afterthoughts 𝗧𝗛𝗘 𝗦𝗜𝗚𝗡𝗔𝗟: Every major cloud provider is building agent-native infra — the tools we built for human developers don't scale to autonomous ones 𝗧𝗛𝗘 𝗤𝗨𝗘𝗦𝗧𝗜𝗢𝗡: How long before your CI/CD pipeline has more agent committers than human ones? If you're running AI coding agents at any scale, the GitHub bottleneck is real. This isn't about replacing GitHub for human workflows — it's about recognizing that agent workflows need purpose-built infrastructure. Anyone else hitting GitHub's walls with agent workloads? Curious what workarounds you've found. Full code + walkthrough → cloudedventures.com #AIAgents #DevOps #CloudEngineering #GitHub #Cloudflare

Here's my developer setup that costs $0 in tooling fees (well, almost). I've been running OpenCode + Oh-My-OpenAgent on WSL + Rancher Desktop, powered by a GitHub Copilot subscription — and honestly, I'm not going back. Here's why this stack hits different: 💳 GitHub Copilot Subscription (the only thing you pay for) - One subscription, access to multiple LLM providers: OpenAI, Anthropic (Claude), Google (Gemini), and more - Switch between models depending on the task — use Claude for reasoning, GPT for speed, Gemini for context - Way cheaper than subscribing to each provider individually - You're not locked into one model — you pick the best tool for the job 🔓 OpenCode - Open-source AI coding agent — no Cursor subscription needed - Terminal-native. Blazing fast. No Electron bloat. - Connects directly to your GitHub Copilot models — plug and play - You own your config, your prompts, your workflow. 🧠 Oh-My-OpenAgent - Supercharges OpenCode with orchestration, delegation, and multi-agent workflows - Skills system lets you load specialized behaviors on demand - Background agents run in parallel — research while you code - All those LLM models from your Copilot sub? Now they're working together as a team. 🐧 WSL (Windows Subsystem for Linux) - Full Linux dev environment without leaving Windows - Native filesystem performance, real shell, real tools - Your team on Mac/Linux? Doesn't matter — same toolchain. 🐮 Rancher Desktop - Docker-compatible, no Docker Desktop license headaches - Kubernetes built-in for local testing - Lightweight, open-source, enterprise-friendly The combo? For the price of a single GitHub Copilot subscription, you get an AI-powered, multi-model, container-ready, Linux-native dev environment running on Windows — with near-zero vendor lock-in. Compare that to: Cursor Pro + ChatGPT Plus + Claude Pro + Docker Desktop seats. Do the math. 💰 If your team is still stacking AI subscriptions, there's a smarter way. #OpenSource #DevTools #AI #WSL #DeveloperExperience #OpenCode #RancherDesktop #GitHubCopilot #LLM

Anyone else follow The Pragmatic Engineer? This mornings email really hit home for me.. One nine. That's GitHub's availability over the past 90 days according to third-party monitoring. Roughly 90% meaning ~2.5 hours of degraded service daily. For the platform 100 million developers depend on. What's driving it? AI coding agents are generating unprecedented volumes of commits, pull requests, and repos. GitHub's CTO acknowledged these outages came during rapid usage growth that exposed scaling limitations in their architecture. Database saturation, cascading failovers, Redis failures, 37 incidents in February alone. At #CircleCI, we've been tracking a related trend. Our 2026 State of Software Delivery report, analyzing 28 million workflows, showed AI increased development activity by 59%, but delivery actually slowed for most teams. We called it the validation bottleneck: code is being written faster than it can be tested, reviewed, and shipped. Now we're seeing the other shoe drop. It's not just validation pipelines under pressure, it's the entire infrastructure layer. And this raises a question every engineering leader should be asking: is your toolchain ready for AI-scale development, or just your code editor? During that same 90-day window, CircleCI maintained 100% uptime across nearly every service category: Machine Jobs, macOS, Windows, UI, API, Runner, Webhooks — with 99.99% on Pipelines & Workflows and 99.98% on Docker Jobs. All while processing over a million builds a day. Our most recent incident? Caused by GitHub's API failing upstream, not our own infrastructure. That's not an accident. It's the result of building infrastructure designed to scale alongside customers, not scrambling to catch up after things break. The AI era is stress-testing every platform in the delivery chain. The teams pulling ahead aren't just the ones writing code faster, they're the ones whose infrastructure can absorb, validate, and ship it without buckling. If your toolchain can't keep pace today, it won't keep pace tomorrow.

Scaling a massive project requires more than just chat interactions; it demands the codification of your engineering DNA directly into the workspace. By embedding GitHub Copilot instructions for security protocols, architectural patterns, and testing standards, you ensure every engineer on the team stays aligned with your core technical vision. This shift transforms AI from a simple autocomplete tool into a persistent guardrail that enforces high-quality software design and rigorous security checks in real time. When your instructions live within the repository, the AI doesn't just write code, it actively defends your architecture and scales your best practices across the entire organization. Try moving your governance into persistent instructions today and watch how it elevates the consistency and security of your entire codebase. #Github #Copilot #Microsoft

For early-stage open-source projects, the “Getting started” guide is often the first real interaction a developer has with the project. If a command fails, an output doesn’t match, or a step is unclear, most users won’t file a bug report, they will just move on. Drasi, a CNCF sandbox project that detects changes in your data and triggers immediate reactions, is supported by our small team of four engineers in Microsoft Azure’s Office of the Chief Technology Officer. We have comprehensive tutorials, but we are shipping code faster than we can manually test them. https://lnkd.in/e9b65Acq