Skip to main content
/connect bridges two areas of your vault that normally live in separate silos. It reads across both domains, identifies shared principles and patterns, and surfaces non-obvious links that you can explore or capture as new thinking notes.

Usage

/connect [domain1] [domain2]
Example: /connect leadership systems-thinking

What it does

1

Scan both domains

Claude finds all vault files related to each domain — notes, daily reflections, research, and context files that touch on either topic.
2

Find bridges

Claude identifies shared themes, patterns, or tensions between the two domains. These are the places where the same underlying principle appears in different forms.
3

Generate novel connections

Claude proposes non-obvious links based on evidence from your vault — connections you wouldn’t have found by reading either domain in isolation.
4

Suggest actions

Claude closes with concrete suggestions: notes to write, experiments to run, or further connections to explore.
The full analysis is saved to thinking/agent-output/ as YYYY-MM-DD-connect-[domain1]-[domain2].md. These skills work well alongside /connect for deeper exploration:
SkillWhat it does
/traceTrack how a single idea evolved over time across your vault
/driftCompare your stated intentions against your actual behavior patterns
/ideas-genGenerate new ideas by scanning across all vault domains for cross-domain patterns
/graduatePromote ideas from daily notes to standalone files in thinking/
Use /connect when you sense there’s something interesting between two areas but can’t quite articulate it. The domains don’t need to be obvious candidates — the less obvious the pairing, the more surprising the connections tend to be.