This NEW Claude Skill Is Blowing People's Minds (Cook)
A walkthrough of Cook, a loop skill that makes Claude keep building against reference pictures and isolated judges until it hits a 99.9% bar.
November 12thCreator
A walkthrough of Cook, a loop skill that makes Claude keep building against reference pictures and isolated judges until it hits a 99.9% bar.
November 12thA 39-minute illustrated walkthrough of the context layer: what it is made of, where each piece should live, how to keep it fresh, and how to run it for a small team.
November 4thA walkthrough of Claude's new query-backed Dashboards feature, and how to wire one to a Supabase database so it stays live instead of freezing like an old artifact.
October 9thA 56-minute walkthrough of rebuilding a Claude Cowork setup around the new cloud-first model, in five layers from context files to scheduled routines.
October 6thA walkthrough of OpenAI's always-on ChatGPT agent, and why skipping the context setup turns it into a confused bot with no idea who you are.
October 1stA 10-minute walkthrough of Anthropic's internal classification of agent loops — four types, two slash commands, and the stop-condition rule that prevents a $6,000 night.
June 30thA 14-minute pyramid framework for AI system design — KISS, need-first, clean data, human gate, model portability — and why the boring stuff is the product.
June 29thAn 18-minute walkthrough of the three MCP harvests — Gmail, Slack, and call recordings — that keep an AI operating system's context from going stale.
June 24thA 13-minute live demo where a security plugin catches 15 out of 15 planted vulnerabilities with zero false positives.
June 23rdAn 11-minute walkthrough of reverse prompting — purpose-built interview skills that extract your tribal knowledge and simultaneously build the AI workflows your business needs.
June 20thA 14-minute tutorial that converts the feeling of being lost into a five-step repeatable system for learning anything with AI.
June 17thA 9-minute rebuttal that reframes a government AI ban as a boring availability outage — and the six rules that make sure it never hurts you again.
June 16thA 10-minute live demo of a Claude skill that reads every connected SaaS system via read-only MCP connectors and returns a visual HTML data map — security flags, PII exposure, and a build-order recommendation included.
June 15thOne prompt, a spec file, three hours unattended, and a $316 API bill — here is what Claude Fable actually built.
June 12thA 14-minute consulting framework for getting any team to adopt AI without mandates, arguments, or forced rollouts.
June 10thA 17-minute systems walkthrough of building a five-stage skill refinement pipeline with a judge AI, a human gate, and a pointed critique of tools that skip both.
June 6thA 16-minute walkthrough of how Anthropic organizes AI skills internally — and how to map that logic to any business.
June 4thA 9-minute field guide to surviving Anthropic's June 15 billing split — and why the builders who panicked built it wrong from the start.
June 2ndA sixteen-minute walkthrough of Claude Code's Dynamic Workflows, where a deterministic script takes over orchestration that the chat window used to carry.
May 29thA 17-minute blueprint for building a six-figure AI consulting funnel starting with a single in-person event.
May 27thHow one developer wired Gmail, Google Calendar, and a bank API into a four-pod Claude Code dashboard that runs every morning and leaves you a tray of pre-researched actions to approve.
May 23rdA 20-minute systems framework for building Claude skills that actually work — starting with the human workflow, not the model.
May 16thA 14-minute reframe that says you've been asking the wrong question — skills live inside agents, not next to them.
May 6thA 15-minute framework teardown dismantling three myths keeping businesses from building reliable AI operating systems.
May 5th