Journal · 2026

Engineering notes,
opinionated takes.

Field reports from the studio: what we built, what we threw away, and what we learned shipping production software for a small number of teams.

No newsletter spam. No SEO mulch. Posts when we have something to say.

Filter 29 posts
cf:edge DEVOPS 8 MIN READ
DevOps2026 · 09 · 01

The $50/mo stack that scales: Cloudflare Workers, D1, and R2.

A detailed teardown of our edge-first serverless architecture for client MVPs. Zero runaway cloud bills, sub-50ms global latency.

8 min read Read
met:loc PROCESS 7 MIN READ
Process2026 · 08 · 31

Measuring AI engineering output: LOC vs verified features shipped.

AI code tools inflate lines of code while creating review bottlenecks. Why traditional metrics fail and what to measure instead.

7 min read Read
dev:box DEVOPS 8 MIN READ
DevOps2026 · 08 · 25

Zero-trust local dev: running production-identical micro-sandboxes.

Stop putting production credentials on developer laptops. How to use local mocks, ephemeral keys, and containerized sandboxes.

8 min read Read
host? STARTUPS 8 MIN READ
Startups2026 · 08 · 20

Should you self-host the model?

Open weights are close enough and inference costs keep falling, so the build-versus-buy question is live again. The answer is usually no, and the exceptions are not the ones people expect.

8 min read Read
03:00 PROCESS 7 MIN READ
Process2026 · 08 · 17

The overnight PR: when agents work while you sleep.

Async agents finish work nobody watched. The hard part is not the running — it is the morning triage, and the fact that trust has to be decided before the work starts.

7 min read Read
LGTM ENGINEERING 8 MIN READ
Engineering2026 · 08 · 14

Who reviews the AI's code?

Half of committed code is now machine-written and carries 23% higher bug density unreviewed. Most review processes were designed for a bottleneck that no longer exists.

8 min read Read
chmod PROCESS 8 MIN READ
Process2026 · 08 · 11

When your agent gets write access.

The moment an agent stops reading and starts acting, every design question changes. A permissions model built on reversibility rather than sensitivity.

8 min read Read
mcp SECURITY 8 MIN READ
Security2026 · 08 · 07

Tool poisoning is the new SQL injection.

Attackers hide instructions in tool metadata the agent reads and the user never sees. Prompt injection rose 340% in 2026, and MCP is the connective tissue in nearly every incident.

8 min read Read
o5:f5 ENGINEERING 9 MIN READ
Engineering2026 · 08 · 04

Opus 5 or Fable 5: the decision tree we actually use.

One is more capable, the other is half the price — and four of the five questions that decide it have nothing to do with capability.

9 min read Read
eval ENGINEERING 8 MIN READ
Engineering2026 · 07 · 31

Your benchmark is an attack surface.

When the thing being measured can read the ruler, evaluation stops being a measurement problem and becomes a security one.

8 min read Read
esc SECURITY 8 MIN READ
Security2026 · 07 · 28

The model escaped the sandbox. The lab found out from someone else.

Two frontier models broke out of an evaluation, chained a zero-day, and reached production infrastructure. The five-day detection gap is the real lesson.

8 min read Read
aug.2 PROCESS 8 MIN READ
Process2026 · 07 · 21

August 2, for teams without a compliance department.

The EU AI Act's high-risk obligations become enforceable on 2 August — and small teams shipping AI hiring, credit, or eligibility features are in scope without knowing it. A plain read on whether it touches you.

8 min read Read
$/tok STARTUPS 8 MIN READ
Startups2026 · 07 · 17

Your AI feature has a gross-margin problem.

AI-first software runs 20–60% margins vs 70–90% for classic SaaS. The model gave your product a cost of goods sold that scales with usage — and per-seat pricing hides it until it's fatal.

8 min read Read
jr>sr STARTUPS 7 MIN READ
Startups2026 · 07 · 14

The junior developer didn't disappear. The job did.

Postings are down ~40% and the story says AI ate the junior engineer. It ate one job, spec-to-code, and left the harder half — the half juniors used to start from.

7 min read Read
n:1 PROCESS 7 MIN READ
Process2026 · 07 · 07

The great toolchain collapse: when one platform eats your whole pipeline.

Research, build, deploy, and growth are collapsing into single agent platforms. How to take the upside without signing away the exit.

7 min read Read
$exec SECURITY 8 MIN READ
Security2026 · 07 · 05

When the agent can move money: the new fintech attack surface.

Agents that can trigger payments and call smart contracts are a new, largely unmapped attack surface. How it differs, and how to defend it.

8 min read Read
deleg AI 8 MIN READ
AI2026 · 07 · 03

From copilots to delegation: what agent engineering actually requires.

Delegation is a different discipline from autocomplete, and most teams are not set up for it. What an agent task actually needs.

8 min read Read
re:bug ENGINEERING 8 MIN READ
Engineering2026 · 07 · 01

The AI debugging tax: what "it wrote it in five minutes" leaves out.

The model writes the code in minutes and the team debugs it for an hour. The productivity story leaves out the verification bill.

8 min read Read
we+ai AI 8 MIN READ
AI2026 · 06 · 29

What an AI-native small team actually looks like in 2026.

Not the headcount-slashing fantasy and not a normal team that pays for a few subscriptions. How the org chart, the day, and the economics actually change.

8 min read Read
v1. STARTUPS 8 MIN READ
Startups2026 · 06 · 20

The three decisions that kill technically sound startups.

The code works. The architecture is clean. The team is good. And the company still fails to scale.

8 min read Read
llm AI 7 MIN READ
AI2026 · 06 · 18

What AI actually changes for a 10-person business. Not the hype version.

The AI conversation is dominated by large enterprises. Here is what actually changes when a small business adopts these tools, and what stays the same.

7 min read Read
pipe DEVOPS 8 MIN READ
DevOps2026 · 06 · 16

The deployment pipeline we run for every client, from day one.

Most teams treat deployment as something you set up when you have time. We treat it as the first thing we build.

8 min read Read
[T] ENGINEERING 9 MIN READ
Engineering2026 · 06 · 13

Go generics in production: the patterns worth reaching for.

We spent a year ignoring Go generics, six months using them wrong, and the last year finding the handful of patterns that actually belong in a production codebase.

9 min read Read
vibes ENGINEERING 8 MIN READ
Engineering2026 · 06 · 04

Why we stopped fighting vibe coding, and what we found.

Three months of AI-assisted coding in production. What held up, what collapsed, and where it changed the calculation.

8 min read Read
agent AI 7 MIN READ
AI2026 · 05 · 28

AI agents in your business: past the chatbot, into the workflow.

Most "AI agents" are just chatbots with extra steps. Here are the patterns that actually work in production.

7 min read Read
lang ENGINEERING 9 MIN READ
Engineering2026 · 05 · 21

Rust or Go: the decision tree we actually use.

Not a language war. A practical rubric built from real projects, with a decision flowchart.

9 min read Read
0day SECURITY 8 MIN READ
Security2026 · 05 · 14

Security is not a feature. The startup attack surface, mapped.

Most startup security failures are known patterns hit by automated tools against teams that assumed it was someone else's problem.

8 min read Read
+debt PROCESS 7 MIN READ
Process2026 · 05 · 07

The real cost of technical debt in a funded startup.

Technical debt is not a code problem. It is a business problem with a compound interest curve.

7 min read Read
Stay in the loop

One email when we publish. Nothing else.

No drips, no funnels, no "engagement campaigns." Just the next post when it lands, about once a month, sometimes less.

Or grab the RSS · Follow on LinkedIn / X