Weizhena/Deep-Research-skills — reverse-engineered prompt

Build me a reusable deep research skill for AI coding assistants like Claude Code, OpenCode, and Codex.

I want to be able to give it a topic, have it create a clear research outline first, let me review and expand the list of things to research or the fields to collect, then run deeper web research item by item. At the end it should turn the collected results into a clean markdown report with a table of contents.

It should work well for things like academic surveys, tool comparisons, competitor research, market trends, due diligence, and technical evaluations. Please include install instructions for the different assistants, including any needed agent files, Python dependency setup, and config steps. Support English and Chinese versions if practical.

Make the workflow feel human controlled, not fully automatic. I should be able to pause after the outline, add more items or fields, then continue. Look up current docs online if you need to.