/mos:research
Run context-aware research that files findings as typed EvidenceClaim graph nodes.
Working
When to use it
Reach for this when you are in the day-to-day work and need to move. Under the hood it runs Hypothesis-Driven Problem Solving.
You do not have to remember it. Describe what is stuck in plain words and Larry routes here for you. The command is the door; the conversation is the key.
How to use it
Type it inside Claude Code:
/mos:research [topic]The [topic] part is optional context. Leave it off and Larry asks for what he needs.
See it run
What it looks like in the terminal:
What to expect
Larry works it through with you, then files the result to room/**/research/* in your room. It becomes part of the record you can build on.
When it finishes, he proposes your next move. You approve, change it, or set it aside. Every choice you make teaches the room what to surface next.
Stuck at any point?
/mos:help shows what each command does, /mos:act lets Larry pick the next move for you, and /mos:suggest-next ranks what to do next. You are never on your own in here.Related commands
/mos:actRun the next move Larry recommends, without typing the /mos: command yourself.
/mos:file-meetingTurn a meeting recording into structured room intelligence.
/mos:pipelineChain methodologies stage-by-stage.
/mos:validateValidate a specific claim or hypothesis against evidence.
/mos:operatorSee or change the conversation operator state.
/mos:feynman-timeline-refreshRefresh the human-readable Timeline section in FEYNMAN.md.