Go Deep
The cold seer. One sharp answer. The thing that changes the altitude of the work.
When to Use
- When you need the single smartest addition to the current plan
- When the conversation is grinding lateral and needs a vertical shift
- When you want one answer, not options
- Pairs with /mimir (mimir runs go-deep cold in parallel with its primed protocol)
The Protocol
Stop. Look at the current work. Ask yourself:
"What's the single smartest, most radically innovative, accretive, useful, and compelling addition I could make to this plan right now?"
Bring one sharp answer, not a list. The thing that changes the altitude of the work.
Rules
- ONE answer. Not two. Not "here are some options." One.
- No preamble. No "let me think about this." Lead with the answer.
- The answer should surprise. If it's obvious, dig deeper.
- Quality over completeness. A half-formed brilliant insight beats a fully formed obvious one.
- If nothing changes the altitude, say so. "We're at the right altitude" is a valid answer.
How It Works in Practice
The operator says "go deep" (or invokes /go-deep). You:
- Assess the current state of the work (what's in front of you)
- Identify what's missing, stuck, or could be 10x better
- Deliver one sharp insight that reframes or elevates the work
- Stop. Let the operator respond.
Portability
This skill works in any AI assistant that can read the current conversation context. No external dependencies. No scripts. No APIs. Pure reasoning protocol.
Origin
Cognitive prosthetics primitive "Seer" (cold variant). Pollock 2026. Named for the shift from lateral thinking to vertical thinking. Go deep, not go wide.