Data & Analytics Skill
Airflow DAG Patterns
Build production Apache Airflow DAGs with best practices for operators, sensors, testing, and…
$15
Supabase Postgres Best Practices
Postgres performance optimization and best practices from Supabase.
Production-grade Postgres discipline for Supabase projects, organized into eight prioritized rule categories from query performance to advanced features. It turns 'measure don't guess' into a workflow: every query gets an EXPLAIN ANALYZE before deploy, every table gets RLS, every connection goes through the pooler. The result is a data layer that stays fast and secure as you scale instead of saturating at twenty users.
$15 one-time
Add to a kit → Prices include 20% VAT. · Forged on real agency work · one-time, no lock-in
- Type Skill
- Category Data & Analytics
- Delivery Email · instant
- License One-time
Inside the run · no black box
See the actual work before you buy it.
The review sequence the skill applies to any Supabase query or schema before it ships, layered like the security stack it protects:
- Security stack first: anon key stays client-side only, service_role never touches browser code, and every table ships RLS-enabled with auth.uid() policies, so even a leaked JWT cannot cross tenant boundaries.
- Connection economics next: production traffic goes through the pooler on port 6543 (transaction mode, pgbouncer=true for ORMs), direct 5432 stays reserved for migrations. Free tier saturates at 60 raw connections; the pooler serves 1000+ serverless requests from a pool of 15.
- EXPLAIN-driven gate: every new query runs EXPLAIN (ANALYZE, BUFFERS) before deploy. A Seq Scan, high cache-miss buffer reads, or a 10x gap between estimated and actual rows each trigger their own fix, from an index to a VACUUM ANALYZE.
- N+1 sweep: any loop calling .select() per row is rewritten as one nested select (orders with items(*)), and pg_stat_statements is checked for the same query pattern repeating hundreds of times, which is the N+1 smell.
- Index and schema pass: composite indexes ordered by selectivity, partial indexes for hot subsets, GIN for full-text and JSONB, and soft delete via deleted_at plus a partial index instead of destructive DELETE.
- Migration safety last: additive changes ship as nullable-add, then batched backfill, then NOT NULL. Destructive DDL waits 7 days behind application-level deprecation, with a PITR snapshot taken before anything irreversible.
One power source. 6 lines out.
supabase-postgres-best-practices · core
core active · 6 lines
- ✓ writing or reviewing sql
Writing or reviewing SQL queries for performance before they ship
- ✓ designing indexes: b-tree
Designing indexes: B-tree, partial, GIN: for the queries that actually run
- ✓ writing rls policies that
Writing RLS policies that survive a leaked JWT
- ✓ configuring pgbouncer/su…
Configuring pgBouncer/Supavisor connection pooling for serverless and edge
- ✓ eliminating n+1 queries
Eliminating N+1 queries with eager loading instead of per-row fetches
- ✓ running zero-downtime mi…
Running zero-downtime migrations on large tables without lock storms
Yours to keep.
Drag time forward. Watch what stays.
Forever
That's what owning means.
The rented stack
ai writing tool: subscription
expired · access lostanalytics suite: subscription
expired · access lostdesign platform: subscription
expired · access lost(nothing left)
Your forge
-
Stop slow queries at the EXPLAIN ANALYZE step instead of in production
license: perpetual -
Survive a leaked token without cross-tenant data exposure
license: perpetual -
Serve 1000+ concurrent requests from a handful of pooled connections
license: perpetual -
Ship column adds and type changes without locking a million-row table
license: perpetual
subscriptions expire · deeds don't
Everything in the box.
Pick a piece up. Watch it work.
Eight prioritized rule categories with CRITICAL-to-LOW impact ratings
6 parts · one working system · ships instantly by email
This wasn't forged for everyone.
- Not for you if you'd rather rent a tool than own one.
- Not for you if you want someone else to run your stack.
- Not for you if you're happy guessing.
Engineers writing or reviewing Supabase Postgres schemas and queries who want performance and security baked in from the first line.
then this was forged for you.Works with
Universal by design: these run in any AI. Delivered in the open Agent Skills + MCP format (native in Claude); ChatGPT, Gemini, Cursor and Copilot adapt the same files their own way.
- Claude Native format
- ChatGPT Adapts via open standards
- Gemini Adapts via open standards
- Cursor Adapts via open standards
- Copilot Adapts via open standards
Catch what's on your mind.
the air is clear. nothing between you and the forge.
catch a spark: the forge will answer
-
We are a tiny project with maybe twenty users. Is this discipline overkill right now?
The opposite: saturating at twenty users is precisely the failure it prevents, and the rules are cheapest to adopt at the first line of schema. Running EXPLAIN ANALYZE before deploy, enabling RLS on every table, and routing connections through the pooler cost minutes now versus a painful retrofit later.
-
Postgres tuning advice is everywhere, which of these rules only matter because I'm running on Supabase?
It is Supabase-specific where it matters: connection-string discipline between ports 5432 and 6543 that prevents 'too many clients' crashes on serverless, RLS policies designed to survive a leaked JWT, locking guidance tied to Supabase realtime, and a gotchas table for the platform's own sharp edges.
-
Does it audit or inspect my live database for me?
No. It is a rulebook for writing and reviewing schemas and queries, eight prioritized categories with before/after code and a nine-point verification checklist. Running live diagnostics with raw SQL against your actual database is what the supabase-operations package does.
-
How is it delivered?
By email right after purchase: ready to run, downloaded instantly, no setup wait.
-
One-time or subscription?
A one-time purchase; no subscription or hidden fees. VAT (20%) is included.
-
Can I get a refund?
As a digital product, it can’t be refunded once downloaded. That’s why we show exactly what’s inside and who it’s for, right here.