Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/jsagir/mindrian-os-plugin/llms.txt

Use this file to discover all available pages before exploring further.

When you cannot tell what you do not know, /mos:map-unknowns plots known, unknown, and unknowable in a Rumsfeld matrix. The unknowable column is usually where the risk lives. The core insight this command is built on: the most dangerous assumptions in a venture are not the ones you know you are guessing — those you watch. The dangerous ones are the claims you are confident about, state as fact, and have never actually tested. Larry cross-references your room entries to find exactly those: high-confidence claims with thin evidence behind them. This command serves the validate-idea job and pairs naturally with any stage where you have accumulated beliefs about the market, the problem, or the solution.

Usage

/mos:map-unknowns
No arguments required. Larry reads the full room context from room/STATE.md and scans existing entries across all sections before opening. Works at any stage of a venture — the earlier you run it, the more time you have to address what it surfaces.
The unknowable quadrant is not a reason to stop. It is a reason to design experiments that reduce the blast radius of the things you cannot confirm before committing.
autonomous_safe: true — this command runs the Knowns and Unknowns Matrix Framework without requiring human confirmation gates. It files the artifact only after asking for your approval.

What happens

1

Context load

Larry reads room/STATE.md and scans the full room — problem-definition, market-analysis, solution-design, and any other filed sections — to build a picture of what the room currently claims.
2

Pace check

Larry asks: Quick pass or deep dive? Quick pass produces a ranked matrix in a single pass. Deep dive works one quadrant at a time — Larry never rushes this exercise, because it is a thinking exercise, not a checklist.
3

Known knowns quadrant

Larry surfaces the claims the room states with high confidence and backs with solid evidence. These are the things you actually know. Larry challenges each one lightly — what’s the source? how recent is it? — before marking it confirmed.
4

Known unknowns quadrant

Larry surfaces the questions the room has explicitly flagged as open: things you know you don’t know yet. These are already on your radar. Larry checks whether any have been informally answered in room entries without being formally closed.
5

Unknown unknowns quadrant

This is the high-value quadrant. Larry cross-references room entries looking for claims stated with confidence — no hedging language, no source citation, no evidence tier marker — and checks whether the underlying evidence actually supports the confidence level. These are the assumptions you didn’t know you were making.
6

Unknowable quadrant

Larry identifies the claims that cannot be resolved with currently accessible evidence — things that depend on future market behaviour, regulatory change, or competitor decisions. For each, Larry notes the minimum viable signal that would reduce the uncertainty.
7

Prioritisation

Larry ranks the unknown unknowns and unknowables by a confidence × evidence-gap score: the higher the confidence and the thinner the evidence, the higher the priority. This gives you an ordered list of the most dangerous beliefs to address first.
8

Artifact creation and filing

Larry composes the unknowns matrix artifact. Before writing, Larry asks: File this to problem-definition? On confirmation, the artifact is filed to the room’s unknowns path.
9

Methodology chaining

If the matrix reveals a pattern — a cluster of unknowns in one domain, a core assumption that underlies many others — Larry names it: Your matrix reveals [pattern]. Want to explore that with [methodology]?

Output / Artifacts

The command produces a Rumsfeld matrix document filed to the room:
room/**/unknowns/<timestamp>.md
The artifact contains:
  • Known knowns — confirmed claims with evidence
  • Known unknowns — open questions the room already tracks
  • Unknown unknowns — high-confidence claims with thin evidence, prioritised by confidence × evidence gap
  • Unknowables — claims that cannot be confirmed yet, with minimum viable signals for each
  • Priority list — the top assumptions to address, ordered by risk weight
The unknown unknowns list from this command feeds directly into /mos:bono as debate axes — the highest-risk assumptions become the what-if hypotheses the six hats argue over.

Example

/mos:map-unknowns
Larry scans the room and opens the matrix. In the unknown unknowns quadrant, Larry surfaces three claims from the market-analysis section stated with high confidence but no source: “Enterprise buyers make procurement decisions in 60 days,” “The incumbent solution has a 70% churn rate,” and “Our target segment has a budget line for this category.” All three were asserted in session notes with no evidence citation. Larry ranks them: the procurement timeline assumption is the highest priority because it underlies the entire go-to-market motion. The others follow. Filed matrix includes a minimum viable signal for each unknowable: one customer discovery call would confirm or kill the procurement timeline claim.

/mos:bono

Run a six-hat governed research debate — the unknown unknowns become what-if hypotheses.

/mos:beautiful-question

Reframe before you map — surface the real problem before auditing your assumptions about it.

/mos:analyze-needs

Ground the unknowns in real customer jobs before deciding which to address first.

/mos:grade

Grade the room’s evidence quality — pairs with the unknowns matrix to close the loop.

Build docs developers (and LLMs) love