The anatomy of a good prompt: how to combine task, context, format, constraints, and examples into requests that get useful answers on the first try.
Five techniques for working prompts harder: role framing, few-shot examples, output schemas, iteration loops, and getting the model to ask you questions before answering.
Techniques for hard, multi-step work: decomposing tasks, using chain-of-thought where it actually helps, self-critique passes, chaining prompts into pipelines, and evaluating results systematically.
A workflow for writing long documents with AI: lock a brief, generate and refine a table of contents, draft chapter by chapter with continuity context, then revise in focused passes.