GitHub Unifies Actions Retention: CI/CD Run Metadata and Checks Align to 90-Day Policy

GitHub announced that starting October 1, 2026, checks, workflow runs, and commit statuses will follow the unified Actions retention policy that currently governs logs and build artifacts. While these execution records were previously preserved for over 400 days regardless of custom repository settings, they will now be pruned when they exceed the configured retention window—defaulting to 90 days across repositories, organizations, and enterprises. The maximum retention for public repositories remains capped at 90 days, and the change will not retroactively restore previously evicted data. This policy shift directly impacts DevOps engineers, security compliance teams, and platform architects who depend on multi-quarter audit trails and deployment verifications. In many regulated industries, compliance frameworks mandate extensive histories of build attestations, automated check approvals, and test results. Relying on GitHub's native UI to inspect a pipeline that ran six months ago will no longer be possible unless organization-level retention is explicitly configured or execution telemetry is piped into external long-term storage. From a broader cloud and DevOps architecture perspective, this update reflects a maturing pattern in SaaS continuous integration platforms: separating real-time workflow orchestration from long-term compliance storage. As developer velocity surges and high-frequency deployment models generate millions of workflow executions daily, maintaining unbounded metadata in primary application databases degrades control plane latency and query responsiveness. Platform vendors are standardizing on aggressive retention windows to preserve system scalability, effectively pushing long-term analytical and compliance storage toward object stores and dedicated data warehouses. In practice, platform teams should immediately evaluate their current retention configurations before the October 1 cutoff. First, review repository and organization-level retention ceilings, extending private repository windows up to organizational policy boundaries if immediate access to recent quarterly runs is necessary. Second, organizations requiring longitudinal build telemetry or compliance traceability should establish automated pipelines to ship GitHub Actions workflow run webhooks, OIDC-signed provenance, and test artifacts directly to external storage like Amazon S3 or Google Cloud Storage. Finally, FinOps teams should note that while metadata pruning may slightly reduce attached log and artifact storage footprints, actively increasing retention settings across entire enterprises will drive higher monthly billable storage costs.