01Confirm the project and table
Compare the client URL, schema, table, filters, and environment with the dashboard you inspected.
What the result tells you: A development-versus-production mismatch can produce a valid empty result.
ERR-749/Supabase · SELECT returns []
Appears when: SELECT returns []
Supabase returns an empty array although data exists
An empty array often means the query succeeded but Row Level Security made every row invisible to the active role. Confirm the project and schema, then compare the session and SELECT policy against the rows you expect to see.
How do I fix supabase returns an empty array although data exists?
01
# Prove the target data existsCheck the exact project, schema, table, and representative row without exposing sensitive data.02
# State the intended visibility ruleDefine which users or tenants should see that row.03
# Correct the policy or ownership dataMake the narrow rule and stored ownership values agree.04
# Test visible and invisible casesVerify the owner can read the row and an unrelated user cannot.Preserve the current working revision, change one proven boundary, and verify the published path in a clean session.
What if the quick fix does not work?
02Inspect the active session and role
Check whether the request is anonymous, authenticated, or server-side and whether the expected user id is present.
What the result tells you: A SELECT policy using auth.uid() will hide rows from an anonymous or different user.
03Evaluate the SELECT policy against a real row
Compare owner and tenant fields to the current claims.
What the result tells you: A policy can be syntactically valid while excluding every existing row.
| Likely cause | What proves it | First safe action |
|---|---|---|
| RLS filters every row | A development-versus-production mismatch can produce a valid empty result. | Prove the target data exists |
| Wrong project or environment | A SELECT policy using auth.uid() will hide rows from an anonymous or different user. | State the intended visibility rule |
| Filter value is different | A policy can be syntactically valid while excluding every existing row. | Correct the policy or ownership data |
| Session has not resolved | A development-versus-production mismatch can produce a valid empty result. | Test visible and invisible cases |
Why do AI-built apps hit this problem?
An empty array often means the query succeeded but Row Level Security made every row invisible to the active role. Confirm the project and schema, then compare the session and SELECT policy against the rows you expect to see.
RLS filters every row
No SELECT policy matches the active role and row ownership.
Wrong project or environment
The app queries an empty production database while the dashboard shows development data.
Filter value is different
Tenant, user, status, or date filters do not match stored values.
When the built-in AI fix makes it worse
Freeze generated changes, restore the last known working version, reproduce one request, collect the browser and platform logs, and change one layer at a time.
How do I stop this problem recurring?
- →Add RLS tests with realistic user and tenant claims.
- →Label development and production projects clearly in configuration and operational tooling.
- →Keep production configuration in an explicit release checklist, not in chat history.
- →Test the published URL in a clean browser session before calling the release complete.
- →Keep a last-known-good deployment and a documented rollback step.
- →Log the critical request path with enough context to identify the failing layer.
Still stuck with supabase returns an empty array although data exists?
AI prototype hardeningWe restore the critical path and document the root cause.
- Production failure reproduced with evidence
- The smallest responsible layer repaired
- Rollback and prevention guidance included
Supabase returns an empty array although data exists questions
Why is there no Supabase error?+
RLS can validly return only rows the current role may see. If none match, an empty array is a successful result.
Does the dashboard prove the browser should see the rows?+
No. Dashboard or service-role access may be privileged. Test with the same role and claims as the browser.
Should I add a policy that returns true for everyone?+
Only if the data is intentionally public. Otherwise encode the actual ownership or tenant rule.