This is how you extract the stuff AI can't fake. It asks 2-4 targeted questions to pull out contrarian takes, specific numbers, project failures, and scope limitations, then packages the answers into a knowledge document you can feed directly into write-content. The questions adapt by content type (how-to vs thought leadership vs case study), and it follows up when answers get interesting. The honest take: this only works if you actually have experience worth extracting. If you're writing about something you haven't done, no amount of good questions will save you. But if you've shipped the project or lived the problem, this turns 10 minutes of conversation into the kind of first-party material that makes content impossible to commoditize.
npx -y skills add inhouseseo/superseo-skills --skill expert-interview --agent claude-codeInstalls into .claude/skills of the current project.
Select a file.
supercent-io/skills-template
supercent-io/skills-template
huangjia2019/claude-code-engineering
remotion-dev/remotion
reactjs/react.dev