Same Local Model, 3 Coding Agents: pi 12/12, Codex 12/12, Claude Code 7/12 on a 6GB GPU
Claude Code told me the web page was done. The folder was empty. That was the worst moment of a sma…
Programming
Software development, code, and open source.
13K articles
The Take-Home Where the Model Is Wrong on Purpose
A candidate shipped his take-home on a Tuesday night. Tests green, README tidy, forty-minute turnar…
Architecture Review: Keep Eval Provenance on the Control Plane, Not the Borrowed Runtime
Last week I ran a four-shard eval batch against a borrowed runtime, went to sleep, and read the led…
I built an offline invoice sorter for freelancers, then measured whether local inference is actually cheaper
This is a submission for the Hacktoberfest Open-Source AI Challenge Week 2: Side Hustle What I Buil…
How a 34 GB log filled my Mac's disk, and how I found it
My Mac mini has a 256 GB disk. On 11 October it had 70 GB free, which is fine until you want Xcode …
How One Edge Trick Slashed Cloud Bills by 87 % and Cut Robot Latency to 1 s
EdgeWise Rolls Out: How Selective Cloud Delegation Shrinks Costs and Keeps Latency Low for Long‑Hor…
Your First Mobile AI PR: Point Inference at a Dev Backend, Then Cut the Network
A new engineer joined our mobile AI repo on a Tuesday, and by 10:40 she had the transcript feature …
Price Free AI Capacity in Deadline Slack, Not in Tokens
It is 02:14. Your nightly batch window closes at 02:30. The free lane is still running, and the tok…
Why Your LLM Agent's "Success" Might Be an Illusion
Your LLM agent just told you it published the document. HTTP 200, all tokens logged, latency normal…
Default-Deny Egress Before Your Agent Meets a Free Model Endpoint
Here is a script I keep watching play out. A team reads that a free tier exists, spins up an agent …
We Tried to Break Our Own SCSS Compiler. Here Are the Numbers.
NuvoUI does something most SCSS frameworks avoid: its apply() mixin reads a list of utility items, …