Published November 15, 2026. A repository can be quiet while the software built from it still runs. That distinction should drive your response to GitHub's upcoming Dependabot change. Below are the announced rules, followed by an original readiness checklist for teams responsible for long-lived applications. Nerova has not audited your repositories or measured the effect of this change.
What changes on December 8, 2026?
GitHub announced on November 10 that Dependabot alerts will be disabled on most repositories inactive for a year, starting December 8. Existing alerts are archived rather than deleted. Activity means a push to any branch or tag, repository creation, or enabling Dependabot alerts.
A push or manually enabling alerts restores the archived alerts and restarts the year. Exemptions include an enforced security configuration enabling alerts, or an enterprise policy preventing repository administrators from toggling them. The change covers github.com plans and Enterprise Cloud with data residency, not Enterprise Server. Read GitHub's announcement.
Inactive code is not necessarily unused software
Our recommendation is to inventory what still runs before classifying repositories as unimportant. A stable service, a scheduled reporting job or a rarely updated internal application may continue to process real data without frequent commits. Conversely, an actively discussed prototype may have no deployed users. Repository activity and operational exposure answer different questions.
Begin with deployment records and accountable owners, rather than assuming the latest commit tells the whole story. Include container images, scheduled jobs and client installations where relevant to your organization. Record unknown deployment status as unknown; an absent owner is a follow-up task, not evidence that a system is retired.
| Inventory field | Question to resolve |
|---|---|
| Repository and owner | Who can verify use and approve remediation? |
| Deployed artifact or service | Which running system depends on this source? |
| Operational exposure | What data, users or critical workflows are involved? |
| Current alert setting and policy | Is monitoring actually enabled and centrally governed? |
| Dependency evidence | Do the checked manifests match the deployed revision? |
| Next review and action | Who will resolve each gap, and when? |
Keep this inventory access-controlled. You do not need to publish repository names, vulnerability details or internal infrastructure to demonstrate progress.
Restoring alerts is not fixing dependencies
Dependabot identifies vulnerable dependencies from supported dependency data and reviewed advisories; it does not detect every security problem. GitHub's documentation describes default-branch scanning and says archived repositories are not scanned. An archived alert in this new policy is not the same thing as an archived repository. Check both states rather than assuming a repository archive preserves active monitoring. GitHub's alert scope and limitations.
For a maintained application, separate three outcomes: monitoring is enabled, findings have been reviewed, and an approved remediation is deployed and verified. A settings change only establishes the first. A merged dependency update does not establish the third if the running service still uses an older artifact.
For each important finding, assign a decision: investigate applicability, prepare an update, accept a documented temporary mitigation, or retire an unused system through the normal change process. Do not let a quiet dashboard substitute for that decision. Keep an owner and an explicit review date on unresolved work.
Choose policy deliberately, not with artificial activity
For an individual repository, GitHub documents the alert control under Settings, then Advanced Security in the Security and quality area. Use an authorized administrator and verify the resulting state. At organization scale, evaluate the available security configuration and enterprise policy controls with the people responsible for them. Official configuration instructions.
We do not recommend a scheduled empty commit solely to keep a dashboard green. That obscures the difference between maintained software and a clock reset. A defensible policy should express which repositories need continuing coverage and who owns exceptions. Confirm the scope before applying a broad setting: changing a whole organization is not equivalent to reviewing one repository.
Review adjacent options before saving a security configuration. GitHub notes that including certain Code Security or Secret Protection features can introduce usage costs or licensing requirements for private and internal repositories. That is a reason to inspect the configuration, not a claim that this announcement makes basic Dependabot alerts a paid feature. Security configuration guidance.
Measure coverage rather than alert volume alone
Use a stable, explicitly defined inventory of in-scope deployed repositories. Track how many have verified monitoring, how many lack an owner, how many have unresolved coverage questions, and how long actionable findings wait for a decision. Keep the check date and evidence for each status.
For a hypothetical example, if 40 repositories are in scope and 30 have verified monitoring, coverage is 75%. If five still need investigation, show those five as unknown, not covered. These are illustrative numbers, not Nerova or GitHub measurements. If the inventory is incomplete, label the result provisional instead of implying company-wide coverage.
A lower alert count can mean fixes, changed dependency data, dismissed findings or lost monitoring. Compare the underlying transitions before presenting it as a security improvement. Likewise, measure a deployed fix separately from a closed ticket or generated pull request.
A practical readiness sequence
- Identify long-lived deployed software and assign repository owners.
- Check actual alert settings, repository archive status and governing policies.
- Resolve unknown coverage with authorized administrators, without broadening access unnecessarily.
- Review important findings against the deployed revision and prepare approved remediation.
- Recheck coverage after the change begins, keeping a record of exceptions and unresolved work.
If you are deciding which maintenance processes to automate, scope the ownership and approval process first. Nerova's operational-improvement project can help frame that planning work. This article does not claim an existing Dependabot integration, an automatic security audit or permission to change your GitHub settings.