→I'll just generate this with Claude / Codex in an hour.
+
You'll generate something that looks like this in an hour. That's
not the same thing. LLMs are confidently wrong about AWS specifics — an IAM
policy that grants sts:AssumeRole
too broadly, a DynamoDB key schema that hot-partitions at 10k writes, a
Cognito trigger that silently drops errors. You won't know until prod.
The hard part of AWS isn't generating any single file — it's getting 30
files across Lambda, API Gateway, DynamoDB, Cognito, Terraform, and GitHub
OIDC to agree with each other. That's not a one-shot prompt; it's 40
prompts, 20 debug loops, and a day of "why does terraform plan
show a diff every run." You become the QA for code you didn't write.
The $149 isn't for the code. It's for decisions already argued and
verified in production — single-table vs. multi-table, JWT vs.
Lambda authorizer, DLQ alarm thresholds, retention policies. An LLM picks
a default. The kit ships with defaults someone actually got paged
for. Treat it as the stable base Claude writes your features on top of, not
the thing Claude replaces.
→Why not use a free GitHub boilerplate?
+
Go for it if you find one that fits. Most are abandoned, pinned to dead
dependency versions, use long-lived AWS keys in CI, or were written to
showcase one feature — not run a real product. These kits are maintained
with 81 passing tests, OIDC CI, and an opinionated path from clone to prod.
You're paying for the parts you don't see: decisions already argued, traps
already sprung, updates for 12 months.
→Who wrote these? Why should I trust them?
+
Mohammed Balila — senior backend engineer, 7+ years
shipping production systems across fintech (Emirates NBD), prop-tech
(Stella Stays), and real-estate (Aqar). AWS Certified Developer
– Associate and HashiCorp Certified: Terraform
Associate. Every pattern in these kits got picked because
something simpler broke at 3am.
Don't trust the pitch — check the receipts:
LinkedIn
·
GitHub
·
email.
ARCHITECTURE.md and DEPLOYMENT.md are one message away.
→What does it cost to actually run on AWS?
+
Cheaper than most people think. ~$0.50/mo idle.
~$5–15/mo at 1k MAU. ~$50–150/mo at 100k MAU. CloudWatch Logs and Cognito
MAUs are the real cost drivers — Lambda + HTTP API stay pennies until
you have genuine traffic. No surprise bills. No min commit.
→Will I get locked in?
+
Not to us. You get the full source zip with git history — no license
server, no phone-home, no "as a service." After purchase, you could
delete this site and your code would still work. You're obviously
"locked in" to AWS itself, but that's what you chose when you came here.
→Do I need AWS experience?
+
Enough to be dangerous — AWS Console, CLI, basic Terraform. Every kit
ships with DEPLOYMENT.md: a walkthrough from a fresh AWS account to
a live dev API. If you can follow a README, you can deploy it.
→License? Can I use it for client work?
+
Yes. One purchase, unlimited projects — your products, client work,
internal tools, across your entire org. The only thing you can't do
is resell or redistribute the template itself. Full terms in the
LICENSE file shipped with every kit.
→Do I get updates?
+
Yes — 12 months of free patch and minor releases. Major versions
discounted. When AWS deprecates something (as it does), you get
the fix — you don't have to find it.
→What if I'm not sure it's the right fit?
+
Read the real docs before you buy. Drop an email to
mohammed@protomindlabs.com
and ARCHITECTURE.md + DEPLOYMENT.md are back within a day. If
after reading them the stack isn't what you need, you'll know
before you spend $149 — and if anything's unclear once you're in,
I answer every email personally.