PromptwireLive Standings

TheWire

Coverage, analysis, and dispatches from the frontier. How the models think, where they break, and what the standings actually mean.

Subscribe · RSS
Filed underField guideClear
Abstract navy and copper illustration of five glowing nodes connected by trade lines, suggesting a small simulated economy.
Field guide

Thousand Token Wood and the case for designing around small models

A Build Small Hackathon project ran a five-agent economy on Qwen2.5-3B and found that scarcity, prompting, and parse-repair did the work people usually assign to bigger models.

By the Promptwire desk
Abstract nested folders and modular blocks floating in deep navy space with copper and teal accents
Field guide

Anthropic's internal playbook for Claude Code skills

Anthropic published a field report on how it uses Claude Code skills internally, sorting hundreds of them into nine categories and laying out the patterns that actually hold up in production.

By the Promptwire desk
Abstract layered blueprint with glowing copper pathways and teal nodes against a deep navy background.
Field guide

Anthropic's playbook for Claude Code in million-line codebases

Anthropic published its observed best practices for running Claude Code across monorepos, legacy systems, and multi-repo enterprises. The headline argument: the harness around the model matters more than the model itself.

By the Promptwire desk
Violet pattern representing reasoning-time spread across the field
Field guide

Reasoning time isn't IQ. But it might be conviction.

Grok wrapped in 84 seconds. ChatGPT took 28 minutes. We unpack what the spread means — and what it doesn't.

Promptwire