Skip to content

A full-featured CRM built with React, shadcn/ui, and Supabase.

Require explicit grants in deploy-time migrations

fzaninottoa05a4b859505
diff --git a/.claude/agents/quality-reviewer.md b/.claude/agents/quality-reviewer.md
index 319e7302..b8a3a3d7 100644
--- a/.claude/agents/quality-reviewer.md
+++ b/.claude/agents/quality-reviewer.md
@@ -146,11 +146,16 @@ Migration checklist (BLOCKING):
- Idempotent (`IF [NOT] EXISTS`), no destructive change without intent.
- Column types/constraints/FKs match the TS types the migration is derived from.
- RLS enabled + real policies on every new table (never `USING (true)`).
+- Explicit `GRANT`s to `anon`/`authenticated`/`service_role` for every new
+ table, view and `serial` sequence, in the migration itself (Supabase no
+ longer auto-exposes new `public` objects; a missing grant passes locally
+ and fails with 42501 in production).
- View-recreation rule respected: `CREATE OR REPLACE VIEW` with the new column
as the LAST item in the SELECT list (after every existing column AND every
existing computed `AS` alias). `DROP VIEW … CASCADE; CREATE VIEW …` only
- for column removal/rename, with dropped dependents re-created in the same
- migration. Column order in `03_views.sql` must mirror the deployed view.
+ for column removal/rename, with the view's `GRANT`s re-declared right after
+ the `CREATE VIEW` and dropped dependents re-created (with their grants) in
+ the same migration. Column order in `03_views.sql` must mirror the deployed view.
- No data loss on existing tables; reversible where feasible.
Files to review are listed in the spawn prompt. Read them in
@@ -170,7 +175,7 @@ Detection: your spawn prompt contains `ROLE: quality-reviewer (SIMPLE mode — s
git -C <WORKTREE_PATH> diff "session-base/$SHORT"..HEAD
```
2. **Apply the scope-relevant rubric only** — SIMPLE diffs are small and schema-focused:
- - **A.6b (schema changes)** — no `supabase/migrations/*.sql` in the diff (off-limits to SIMPLE); schema files in `supabase/schemas/*.sql` only; new column appended at the end of the `03_views.sql` SELECT, no ordinal shift.
+ - **A.6b (schema changes)** — no `supabase/migrations/*.sql` in the diff (off-limits to SIMPLE); schema files in `supabase/schemas/*.sql` only; new column appended at the end of the `03_views.sql` SELECT, no ordinal shift; every new table, view and `serial` sequence declared in `06_grants.sql` (no default privileges, so a missing grant means `permission denied`).
- **B.1 (RLS)** — RLS enabled, policies cover required ops, no `USING (true)`.
- **B.3 (injection)** — no string-concatenated SQL, no `||` of user input.
- **A.6 (backend patterns)** — input validation, no unbounded queries.
diff --git a/.claude/skills/writing-migrations/SKILL.md b/.claude/skills/writing-migrations/SKILL.md
index 80bf8e44..d444085e 100644
--- a/.claude/skills/writing-migrations/SKILL.md
+++ b/.claude/skills/writing-migrations/SKILL.md
@@ -100,8 +100,15 @@ Write to `supabase/migrations/<YYYYMMDDHHMMSS>_<SESSION_SHORT_ID>_migration_<slu
correct column types matching the TS types, FKs, and RLS for new tables (RLS
enabled + policies, never `USING (true)`).
+Every new table, view and `serial` sequence needs explicit `GRANT`s to `anon`,
+`authenticated` and `service_role` in the same migration, mirroring the
+declarations in `supabase/schemas/06_grants.sql`. Supabase no longer exposes new
+`public` objects to the Data API automatically, and the schema declares no
+default privileges: without the grants the object answers `permission denied`
+(42501) on a fresh or post-2026-10-30 project, even if it works locally.
+
**Checkpoint:** re-running the migration on an already-migrated database is a
-no-op (every statement is guarded), and every new table has RLS enabled with at least one non-`USING (true)` policy.
+no-op (every statement is guarded), every new table has RLS enabled with at least one non-`USING (true)` policy, and every new table, view and `serial` sequence has its `GRANT`s in the migration and in `06_grants.sql`.
### 5. View-recreation rule (BLOCKING correctness)
@@ -118,12 +125,13 @@ exceptions. The mechanical rule is: positions 1..N of the new view must be
identical to positions 1..N of the old view; the new column is position N+1.
Why not `DROP VIEW … CASCADE; CREATE VIEW …`? It silently drops dependent
-views/materialized views/rules, loses any explicit `REVOKE`s on the view
-(re-granted by default privileges at the next deploy without warning), and
-reads as a destructive operation in audit. Reserve it for cases `CREATE OR
-REPLACE` genuinely can't handle (column **removal** or **rename**). If you
-use it, enumerate dropped dependents in a SQL comment and re-create them in
-the same migration.
+views/materialized views/rules, drops every grant on the view (nothing
+re-grants it: the rebuilt view answers `permission denied`), and reads as a
+destructive operation in audit. Reserve it for cases `CREATE OR REPLACE`
+genuinely can't handle (column **removal** or **rename**). If you use it,
+re-declare the view's `GRANT`s (from `06_grants.sql`) right after the
+`CREATE VIEW`, enumerate dropped dependents in a SQL comment and re-create them
+(with their grants) in the same migration.
PostgREST queries the view, not the table — a missing update makes the
column invisible to the app. The view's column order doesn't matter to API