Corpus distillation -- pick one high-value stable corpus, generate a synthesis study file (narrative, current state, key decisions, anticipated Q&A with citations) into memory/distilled/ so retrieval can answer multi-hop questions that chunk retrieval misses.
$npx -y skills add bwelker/edwin --skill self-studyInstalls into the current project.
Run `npx skills use "https://github.com/bwelker/edwin" --skill "bwelker/edwin/self-study"` and follow the generated skill instructions now. Read its complete output, redirecting it to a temporary file first if necessary. Resolve relative paths from the supporting-files directory it provides.
Use the skills in "https://github.com/bwelker/edwin" that are relevant to the current task. Run `npx skills add "https://github.com/bwelker/edwin"` and select the relevant skills, then follow their instructions.