This week, GitHub described repository access management and Cloudflare introduced a feature for processing network requests. Teams adopting either approach need a recovery plan: someone must be able to investigate and reverse a change that blocks legitimate work.
What GitHub and Cloudflare announced
GitHub's Evergreen account describes recurring collaborator reviews and archival of inactive repositories. Reviews use deadlines and tracking issues; archival preserves the repository. The article describes GitHub's own operating process.
Cloudflare's Gateway announcement introduces a Forward action that routes allowed HTTP traffic through a Worker. At the time of the announcement, Forward was in public beta and Gateway Custom Functions were planned for a later release.
Plan for legitimate work being blocked
An access rule can follow its configuration correctly and still interrupt legitimate work. A request transformation can work as intended and still break an application. Plan how to investigate and reverse those changes when they interrupt work.
For repository access, name the person who can confirm a collaborator still needs access and the person who can restore it. For traffic rules, name the person who can disable a bad rule. In both cases, keep a record of what changed and why so the team can investigate a problem.
Check that another teammate can restore access
Choose one task in a test environment and simulate a legitimate request being blocked. Ask someone other than the person who configured the rule to find the cause in the logs and restore access. Check which permissions and instructions they need, then document any missing steps.
This commentary is based on the linked announcements. The article selection below and this commentary are updated separately. See our editorial policy for correction requests.