2 months ago
Hello team,
I'm facing an issue with Railway's autodeploy feature.
Even though Railway is properly installed and authorized on my GitHub account, automatic deployments are not being triggered when I push changes to the connected repository.
Here’s what I’ve already validated:
Railway GitHub app is installed and has access to the correct repository
The project is properly linked to the repo
The target branch is correctly configured
Manual deploys work without issues
However, pushes to the branch are not triggering any new deployments.
Could you help me understand what might be misconfigured or if there’s any known issue affecting autodeploy?
Thanks in advance.
1 Replies
2 months ago
This thread has been opened as a public bounty so the community can help solve it. The thread and any further activity are now visible to everyone.
Status changed to Open Railway • about 2 months ago
2 months ago
Since the obvious stuff checks out, run through the documented troubleshooting list — in my experience it's almost always #1 or #2:
Check for skipped deployments. In the service's Deployments tab, click "Show Skipped." If your pushes show up as SKIPPED, autodeploy is firing and something is filtering it out:
Watch paths — if configured, any commit that doesn't touch a matching path is skipped. This is the most common silent culprit, especially in monorepos.
Wait for CI — if enabled and your GitHub Actions workflow fails (or doesn't run on push for that branch), the deployment is skipped.
Commit message contains [skip ci] or similar.
Verify the webhook actually reaches Railway. In GitHub: repo → Settings → Webhooks (or the Railway app's installation page) → check Recent Deliveries for push events and whether they got 200s. If deliveries are failing or absent, uninstall and reinstall the Railway GitHub App.
Confirm the branch match is exact — e.g. main vs master, or pushing to a branch that merges into the trigger branch via PR (the merge commit triggers, feature-branch pushes don't).
Autodeploy requires at least one project member with a connected GitHub account that has contributor access to the repo — if the repo moved orgs or your GitHub token was revoked, this breaks silently (requirements).
Quick reset that fixes stale cache cases: disconnect and reconnect the repo in service settings.
If none of that surfaces it, the docs ask you to give support the repo name, branch, and the specific commit SHAs + UTC timestamps that didn't deploy — that lets staff check webhook delivery on their side.