GitHub's status page no longer provides aggregate uptime numbers, which has left users wanting more clarity on service reliability. This article fills that gap by reconstructing the platform’s uptime metrics over the last 90 days. It draws data from archived status updates, breaking down uptime into minute-level intervals and mapping incidents to specific services. This approach gives a more granular view of service performance compared to the previous day-level summaries. The project is open source, encouraging contributions from the community. It utilizes a pipeline that replays Atom feed snapshots stored in Git. By merging downtime windows instead of lumping data into daily buckets, it offers a detailed incident timeline. When certain components are missing, the system infers tags using a tool called GLiNER2, but it only retains labels that are directly referenced in the incident reports. For users relying on GitHub's services, this resource provides a clearer picture of uptime trends and incidents. The article also highlights a dedicated archive that tracks uptime data back to 2017, making it easier to analyze long-term service reliability.