You Approved It and the Agent Asked Again: Approval Is a State Machine, Not a Button
A human-in-the-loop approval that gets recorded, serialised, restored and then silently ignored is the most expensive bug shape in an agent system, because nothing errors. The same defect has been re-found at four different boundaries in one SDK this year. The fix is architectural: stop inferring which call an approval belongs to and write it down.
Read More

Same Weights, Different Answers: We Stopped Signing Off Inference Upgrades on Latency
Ninety catalogued bugs across two open-source serving frameworks, and the single most common symptom was not a crash — it was an accuracy regression. Changing only the inference backend, with the weights untouched, has been measured moving a benchmark by 16.6 points. Here is the gate that replaced our latency dashboard.
17 Nov 2026 · 8 min read
The Failover Worked and the Answers Got Worse: A Model Handoff Is a Quality Event
Routers, fallbacks and forced upgrades mean a second model regularly finishes a conversation the first one started. Benchmarking each model alone never tests that handoff — and the drift it causes is directional, measurable, and invisible on your uptime dashboard.
16 Nov 2026 · 8 min read
Forty-Three Calls to One Tool, Zero Errors: Your Agent Loop Is a Tool Design Bug
Error-rate monitoring asks whether a step failed. Inside a no-progress loop every step succeeds, so nothing fires. The signal you actually need is whether the world changed — and it belongs in the tool layer, not the prompt.
15 Nov 2026 · 10 min read
The Default Value Now Returns 400: An SDK Major Is an API Contract Change
Your dependency PR says one package went from 0.x to 1.0. Nothing in that diff says 2026-10-21, but that is what shipped. The API version your requests use is pinned to your SDK's major, and it is the one field nobody reviews.
14 Nov 2026 · 10 min read
Your Helper on 127.0.0.1 Now Needs Permission: Four Browser-to-Device Architectures, Compared
A printer, a scale, a POS terminal, a local agent on loopback — the integration that has worked for years now depends on a permission prompt the user can dismiss, and the browser reports the dismissal as a failed fetch. Four ways to reach local hardware, ranked by what each does when the answer is no.
13 Nov 2026 · 11 min read
Every Request Returned 200 and the Bill Tripled: Prompt Cache Breakpoints Do Not Survive Translation
The expensive incidents are not the ones that page you. A gateway change that is functionally a no-op can invalidate every cache breakpoint in your prompt, and nothing in a normal monitoring stack will go red: same output, same status codes, same token counts, two to three times the bill.
12 Nov 2026 · 11 min readWork With Us
Let’s create
with purpose
Share your goals, timeline, and challenges — we’ll respond with clarity and next steps.
Ambitious ideas deserve thoughtful execution. Start the conversation and let’s define what success looks like.
Team
Acetrum
Est. 2015
4.9/5
Trusted by
top brands
Newsletter
Signals worth
paying attention
No recycled headlines — just the patterns we’re seeing across real client work, distilled into one read a month.
A curated digest of practical thinking and real-world brand perspectives monthly.
No spam. Unsubscribe anytime.
























