If you got an error message from the orchestrator, search this page
(Cmd+F) for any unique phrase from it.
”API has not been used in project …”
GCP API not enabled. See Cloud 403 errors §1.
”AccessDenied — User is not authorized to perform sts:AssumeRole”
AWS role trust policy wrong. See Cloud 403 §3.
”AccessDenied — sts:ExternalId mismatch”
External ID in form doesn’t match role’s trust policy. See Cloud 403 §3.
”AADSTS7000215: Invalid client secret”
Azure client secret rotated/expired. Re-generate per Connect Azure Step 2.
”Couldn’t load notifications”
Backend / network error on the notifications API. Click Retry. If
persists, check https://status.hivedeploy.in.
”credential probe failed: …”
The Connect button’s probe failed. The full error text after the colon tells you which step. See Cloud 403 errors.
”Email verification link expired”
Verification tokens are valid 24h. Click “Resend verification email” on the login page.
”Mark all read” disabled / not clickable
You have zero unread notifications. The button correctly disables.
”No notifications match these filters”
The filter chip combination has no results. Click “Clear filters” to see all notifications.
”Org member limit reached — upgrade to invite more”
You hit the plan’s seat cap. See Billing.
”Please complete the security challenge”
Cloudflare Turnstile widget didn’t load or you didn’t tick the checkbox. See the Turnstile-specific section in the platform admin docs (internal).
”Preparing your workspace…” stuck
Your JWT lacks an org_id claim; the orchestrator is fetching one.
Should resolve in 1-3 seconds. If stuck > 30 seconds, hard-reload
the page.
”Probe failed: 403”
See Cloud 403 errors.
”Probe failed: Cloud Resource Manager API has not been used…”
See Cloud 403 §1.
”Service account … exists but our principal lacks workloadIdentityUser”
WIF binding mismatch or required GCP API not enabled. See Cloud 403 §1 and §2.
”STRIPE_MOCK_MODE must NOT be true in production”
Self-hosted operator error: STRIPE_MOCK_MODE=true was set in a
production env. Set to false (or remove). See
env-vars.
”Tests are failing” (you’re a developer trying to push)
This isn’t a customer-facing error; it’s a CI failure. Check the build logs for the specific failing test.
”Token expired — please log in again”
Your auth JWT expired (12h TTL). Log in again. The orchestrator doesn’t auto-extend — if you want longer sessions, contact support.
”WIF binding scoped to a different principal”
Your binding’s attribute.org_id/X doesn’t match the orchestrator’s
JWT. See Cloud 403 §2.
”WIF provider doesn’t trust this orchestrator’s issuer”
Your provider’s --issuer-uri doesn’t match the orchestrator’s
actual issuer URL (https://backend.hivedeploy.in). Delete and
recreate the OIDC provider with the correct issuer URI per
Connect GCP.
”You’re all caught up”
Not an error — you have zero notifications.
How to file a bug
If your error isn’t in this list, file at
support@hivedeploy.in with:
- Exact error message
- Steps to reproduce
- Your org ID (from the GCP connect form’s banner or DevTools)
- Timestamp (so we can find the matching audit log entry)
We add new entries to this list as customers hit them.
See also
- Cloud connect 403 errors
- Deployment stuck
- Concepts — Architecture — useful for understanding what subsystem your error comes from