Mine this project's Claude Code session logs into durable improvements — proposed memories, CLAUDE.md rules, docs, skills, and decision-log (DECISIONS.md) entries. Async "AI dreaming": distill raw sessions into friction-focused digests, then consolidate across sessions (dedup vs what's already captured, including any project DECISIONS.md, frequency-gated promotion).
$npx -y skills add devonjones/devon-claude-skills --skill dreamInstalls into the current project.
Run `npx skills use "https://github.com/devonjones/devon-claude-skills" --skill "devonjones/devon-claude-skills/dream"` 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/devonjones/devon-claude-skills" that are relevant to the current task. Run `npx skills add "https://github.com/devonjones/devon-claude-skills"` and select the relevant skills, then follow their instructions.