Most "I built an AI agent" posts come with a screenshot and a vibe. This page is the receipts — live counts from my running stack, refreshed each morning at 06:15 ET. No render-on-load tricks. The system actually does this work.
Cost of running
cloud-API cost · lifetime
Inference runs on the workstation. Cloud only when local genuinely can’t carry the work.
est. monthly savings
vs. Buffer Pro + Hootsuite Pro + ChatGPT Plus + ~$150 cloud-API usage.
posts shipped · last 7 days
Drafted by NEXUS, reviewed by me, fired on weekly slots. Receipts below.
Last 7 days
Agents
Autonomous workers running in my stack — each with a single job, on its own schedule, self-healing, around the clock.
Recently shipped
The most recent posts NEXUS drafted, I reviewed, and the stack published.
Alt + Space has opened the Windows system menu since Windows 3.0. Google just took it for Gemini. The model isn't the product, the keystroke is: read access to your screen, Gmail and Drive the second you get stuck. Th...
Read on bluesky →Two numbers from the same safety checkpoint. Harmful responses: 26.26% down to 0.14%. Over-refusal on a benign test set: 2.00% up to 74.00%. The first number is the result. The second is what it cost, and it is three...
Read on twitter →"Same price as the last version" is a claim about a rate, not about your bill. Google released Gemini 3.8 Flash this week, its third Flash model in six weeks, at $0.75 per million input tokens and $3.75 per million o...
Read on linkedin →Google shipped agentic video understanding this week. One config value, processing set to "agentic", and the model stops ingesting your video at a fixed frame rate. It chooses instead: which segments to load, at what ...
Read on twitter →About 1,050 Japanese municipalities and 550,000 public employees now run on one AI platform. The headline result is 3-5x. That number measures how fast the vendor's engineers ship code. The gain is real and I would n...
Read on linkedin →This page is a static snapshot, pushed once each morning by the same stack that runs everything else. No backend · no live API · no telemetry from your browser. The Sunday newsletter tells the story behind the numbers.