8 posts found
TypeSafe's Jev returns decisions, not text. How it works, what it costs, which benchmark claims hold up, and use cases from healthcare triage to logistics.
Read more →My Claude Code setup: layered CLAUDE.md rule files, a lazy senior dev mode called ponytail, a caveman mode that saves tokens, memory plugins and skills.
Read more →Claude wrote me a clean, accurate, SEO-ready post in twenty minutes, and I deleted it. On AI slop, provenance, and the test for whether a post should exist.
Read more →Portfolio chatbot v2 on a Cloudflare Worker and Astro streams replies, keeps chat memory and renders Markdown links. Plus two bugs found only on the live site.
Read more →What actually clicks each day when you start using Claude Code — install, context, CLAUDE.md, plan mode, skills, git guardrails. A practical 7-day onboarding.
Read more →A portfolio AI chatbot on Cloudflare Workers AI, with KV rate limiting, D1 logging and Astro. No API keys and no per-request fees. A walkthrough of the code.
Read more →AI agent skills for Claude and other models. How to build, structure, test and ship reusable skills, with templates and a step-by-step guide to a first skill.
Read more →How to build AI agents that reason, plan and run multi-step tasks: architecture, tool and memory patterns, common pitfalls, evaluation and a working example.
Read more →