How pgsql-test isolates Postgres tests: a fresh database per test file, a rollback per test, and shared fixtures that survive every rollback. Real Postgres, real migrations, real roles. No mocks. 🧵
Not on vanilla Postgres setups? The same harness powers supabase-test, drizzle-orm-test, pglite-test and graphile-test for PostGraphile/GraphQL. Same isolation, same real-roles testing.
1
27
That's also why agents can rerun the suite all day. Resets are cheap, and there's no mock to fake a pass against. Works with Jest, Mocha or any async runner.
npmjs.com/package/pgsql-test
9
Safegres made this week's @heavybit DevToolsDigest, alongside AWS 🙌
Tens of thousands of AI-built apps hand their data to anyone who asks. Safegres connects read-only and shows what each role can actually reach.
heavybit.com/devtoolsdigest/…
Constructive just crossed 300 million npm downloads — the culmination of a decade of open source work.
301.6M total · 58.9M/month · 16.1M/week
Cloud: 201.1M
Last 30 days: 14.7x vs a year ago. Cloud: 27.4x. 71% of downloads came in the last 12 months.
Most Postgres data leaks aren't clever attacks. They're misconfigurations, and misconfigurations can be checked.
Safegres audits grants + RLS read-only and grades every change in CI. Open source, MIT.
constructive.io/blog/introdu…
2
173
Safegres is live — open-source verifier for Postgres grants + RLS.
Catch open tables before agents ship them. MIT. Runs in CI.
1/6
Hacktober made it clear: security is impossible to ignore.
A wave of disclosures and hacks — not one incident, a pile of them. Agents can write schema all day; the hard part is still who can reach the data.
1
1
219
24,664 apps sending database rows to anyone who loads the page. Row-level security fixes this. A policy on every table decides which rows each user can see, and Postgres checks it on every query, from every client, including agents. Start from default-deny: a table with no policy returns nothing. Security should be the starting state, not a checklist item.
A security firm scanned 159,041 Lovable-built apps:
– 24,664 sent DB records on page load, no login
– 7,403 exposed tables with emails
– Worst: 1,000+ passport records, 200,000+ chats
AI ships the app in an hour. Row-level security still needs an owner. Who checks yours?
3
5
319
The best decision in Postgres history was designing it from day one to be extensible, and the second best was keeping it truly community-owned. Four decades later, almost every "Postgres killer" ended up becoming a Postgres extension 🐘
Postgres is a go-to choice for a lot of teams now, but it hasn't always been. For almost 40 years, it fought off competition and near-death.
What was its defining moment on the way to inevitability? Let's take a look.
planetscale.com/blog/what-wa…
Really glad to see the industry converging on this.
Back in May, we wrote "From Vibe Coding to Agentic Engineering" and got on stage at SVB with engineers from Google DeepMind and Mastra to make this exact case:
"You're not just writing code — you're designing the environment that code lives in... letting agents operate within that environment, but always within constraints that make the outcome trustworthy."
Whether you call it an Agent Habitat or Agentic Engineering, the shift is the same: when code generation approaches zero cost, trust becomes the bottleneck.
One piece we'd add: the boundary can't just be a sandbox around the agent. Authorized agents still touch data—so permissions and Row-Level Security have to be compiled deterministically into the database core itself.
Vibe code the edges. Trust the core.
📝 Our essay: constructive.io/blog/from-vi…
🎥 Full May 1 SVB panel (just dropped today): youtube.com/watch?v=YG_yZ_Q_…
www.teapawt.org/amasad/status/21147691…
Software ate the world, now it's learning to run it. View article
As AI drives the cost of writing code toward zero, the bottleneck doesn't disappear. It moves to trust.
You could vibe code the layout of the front of a bank, but you probably wouldn't want the vault downstairs to be vibe coded.
Full recording is now up from our panel at @SiliconVlyBank with Sanjeev Dhanda (@GoogleDeepMind), @abhiaiyer (@mastra), and @Kristopherfloyd (@frontiersynd) on the shift from vibe coding to agentic engineering.
Full recording is live: Is Software Dead? From Vibe Coding to Agentic Engineering
Sanjeev Dhanda (@GoogleDeepMind), @abhiaiyer (@mastra), and @danlynch (@constructive_io) join @Kristopherfloyd (@frontiersynd) at @SiliconVlyBank to break down what happens when writing code is free and trust becomes the scarce resource.
Highlights inside:
- 04:52 — The microwave metaphor & frontend velocity
- 09:20 — The bank lobby vs. the bank vault
- 20:03 — Tests as a guidance system for agents
- 26:53 — Compiling Row-Level Security into Postgres by default
Watch the full session: youtube.com/watch?v=YG_yZ_Q_…
Our CEO @danlynch opened FRC #26 with Secure-by-Default Postgres for Agents.
Agents are database users now. Authorization has to live in the database, not the app.
Thanks @frontiersynd @Kristopherfloyd @pebble_bed @danielchalef @zep_ai @shulynnliu @shagarw21
Hundreds of speculative queries a second, from a user that isn't human. That's the workload. Authorization has to live in the database, not the app. Proud to sponsor FRC #26 with @frontiersynd @pebble_bed www.teapawt.org/frontiersynd/status/21…
Your next database user won't be a human. It'll be an agent firing off hundreds of queries a second.
@shulynnliu and @shagarw21 of @BerkeleySky made the case at FRC #26 at @pebble_bed: agents don't query, they speculate. Schemas, samples, partial plans, retries, all in parallel.
On BIRD, more attempts meant better answers. Across 50 parallel attempts, most of the work was redundant. Agents already create ~20x more branches and ~50x more rollbacks than humans.
So the database has to change. Probes instead of queries. Satisficers instead of optimizers. Branching storage built for agents.
Thanks to @constructive_io for sponsoring.
3
167













