My GitHub webhook auto-deploys are still not working after the GitHub app installation token incident (incident k5z4d1v1tqmt). My service did not recover automatically. The webhook is completely missing from my GitHub repo — no webhook appears at github.com/zukor/seo-optimizer/settings/hooks. I've disconnected and reconnected the source repo multiple times with no effect.

Please reset the cached GitHub installation ID for my service:

Project: cozy-purpose

Service: Zukor-Optimizer (ID: 90b86990-842e-47e9-98c8-8c09182f7081)

Repository: zukor/seo-optimizer

Environment: production

The service is currently online on the last good deploy — only new auto-deploys are blocked. CLI deploys work fine.

Railway’s GitHub integration does not always show up as a normal repo webhook in GitHub settings, so the missing webhook may not be the actual problem.

I’d check these in order:

1. In GitHub, go to Settings → GitHub Apps → Railway → Configure and confirm the app has access to zukor/seo-optimizer.
2. If GitHub shows “pending permission updates” for Railway, accept them.
3. In Railway, disconnect/reconnect the repo from the service by selecting it again from GitHub, not by manually adding a webhook.
4. Confirm the service has Auto Deploy enabled for the exact branch you push to.
5. In Deployments, enable “Show Skipped” and check whether Railway is receiving the push but skipping it because of branch, root directory, watch paths, or CI settings.

If Railway sees skipped deployments, the GitHub connection is working and the fix is in the service deploy trigger settings. If Railway sees nothing at all after reinstalling/reconfiguring the GitHub App for that repo, Railway staff may need to refresh the repo installation on their side.