GitHub powers the work of millions of developers worldwide, and with that scale comes a serious responsibility to keep the platform reliable and transparent. Earlier this year, the team began addressing availability concerns head-on, and now they're rolling out a trio of major improvements to how they communicate service health. These changes focus on three core principles: transparency, accuracy, and timeliness. In this listicle, we break down the six most important updates you need to understand—from a new incident severity level to per-service uptime percentages and more granular insights for Copilot AI services. Whether you're a developer relying on GitHub for daily work or an administrator monitoring platform health, these enhancements will give you a clearer picture of what's happening when things go wrong.
1. A New Incident Severity Level: Degraded Performance
GitHub has introduced a fresh severity classification called Degraded Performance to better reflect the range of issues that can affect its services. Previously, even minor slowdowns or intermittent errors were labeled as Partial Outages, which falsely implied a service was largely unavailable. This new tier sits alongside the existing Partial Outage and Major Outage states, forming a three-level system that more accurately describes user impact. Under Degraded Performance, the service remains operational but may experience elevated latency, reduced functionality, or sporadic errors affecting a small percentage of requests. By reserving the "outage" labels for true outages, GitHub ensures that developers can quickly judge whether a service is still usable or whether they should switch to a backup plan.
2. A Three-Tier System for More Accurate Classification
Building on the new Degraded Performance level, GitHub now uses a three-tier incident classification framework that provides a nuanced view of service health. The three states are: Degraded Performance (service is impaired but functional), Partial Outage (a significant portion of the service is unavailable or severely impacted for a meaningful number of users), and Major Outage (broad unavailability affecting most or all users). This structure replaces the old binary or limited approach, where any disruption—no matter how small—was lumped into the Partial Outage bucket. The change means that when you see a "Degraded Performance" notice, you know the service is still working, just not perfectly. This clarity helps teams decide whether to wait for a fix, use workarounds, or escalate internally.
3. Per-Service Uptime Percentages Now on the Status Page
Transparency goes a step further with the addition of per-service uptime percentages directly on the status page. For each major GitHub service, you can now see the uptime over the last 90 days at a glance. These percentages are calculated using industry-standard methods that account for incident severity and duration. Instead of a single overall number, you get a granular reliability track record for services like Actions, Packages, Pages, and more. This data empowers users to assess which parts of the platform are most stable and to make informed decisions about where to invest their workflows. It also holds GitHub accountable, as any dip in uptime is publicly visible and easily verifiable.
4. How Downtime Is Weighted for Accurate Uptime Numbers
The uptime percentages are not a simple minutes-of-downtime-to-total ratio. Each severity level carries a specific downtime weight to reflect the real impact on users. For a Major Outage, the entire duration counts as downtime (100%). For a Partial Outage, only 30% of the duration is counted—because the service is still partially available. And for Degraded Performance, the weight is 0%, meaning it does not reduce the uptime percentage at all, since the service remains functional. For example, if a service experiences a 1-hour Partial Outage over 90 days, that counts as just 18 minutes of effective downtime. This weighting prevents minor hiccups from over-penalizing the uptime number while still recognizing that some outages impact more users than others.
5. Granular Insights for Copilot AI Model Providers
One of the most requested features is now live: a dedicated "Copilot AI Model Providers" component on the status page. This component provides clear and separate communication when the AI models powering GitHub Copilot experience availability issues. Previously, users only saw a generic status for the overall Copilot service, which could hide provider-specific problems. Now, if a particular AI model provider (e.g., OpenAI) is having an outage, that status is shown explicitly under this new component. This granularity helps developers understand whether Copilot's functionality is degraded due to an upstream model issue rather than a GitHub infrastructure problem, allowing them to plan accordingly.
6. A Commitment to Transparent, Accurate, and Timely Communication
All these updates are guided by three core principles: transparency (showing the full picture of service health), accuracy (using appropriate severity levels and weighted calculations), and timeliness (publishing updates as soon as possible during incidents). GitHub has also improved its incident post-mortems and communication templates to explain not just what happened, but why and what steps are being taken to prevent recurrence. By making these data points publicly available—like the per-service uptime and the new severity level—GitHub is setting a higher bar for platform reliability reporting. Developers can now hold the company accountable with concrete numbers and clear incident categories, fostering greater trust in the services they depend on daily.
Conclusion
GitHub’s status page overhaul represents a significant step toward more honest and useful communication during service disruptions. The addition of the Degraded Performance level, per-service uptime metrics, weighted downtime calculations, and a dedicated Copilot component all work together to give developers the precise information they need. Whether you’re troubleshooting an issue or simply monitoring platform health, these updates help you distinguish between minor glitches and serious outages. As GitHub continues to invest in reliability, these transparency measures ensure that you’re never left in the dark—and that the platform’s performance is measurable and publicly verifiable.