For the complete documentation index, see llms.txt. This page is also available as Markdown.

Overview

Forward-looking, multi-agent scenario analysis built on Clarity.

Foresight is experimental. It's gated per organization and may change.

Where Clarity answers "what do we know now", Foresight asks "what happens next". It runs multi-agent analysis on top of the same evidence graph to produce forecasts and track how views evolve. It has two surfaces: Simulation and Motif World.


Simulation

Simulation runs multi-agent, multi-round analysis on a market question or hypothesis. Instead of a single answer, it coordinates several specialized analysts, tracks where they agree and disagree across rounds, and produces a synthesized forecast.

A typical run:

  1. Create a simulation and review its estimated cost.

  2. Launch it and follow progress in the live view.

  3. Inspect the synthesis, forecast bands, and per-agent output.

  4. Revisit the run later from history.

Simulation is cost-controlled at the organization level and requires the module to be enabled and budget to be available.


Motif World

Motif World is the map-based, continuously evolving counterpart. Rather than a single hypothesis, it maintains a world state you can watch unfold: events appear over time, agents react, consensus shifts, and you can replay historical windows.

Element
Purpose

World canvas

Map-based view of events, entities, and relationships

Session controls

Start, pause, resume, and replay sessions

Agent debate panel

How agents interpret major events

Timeline controls

Scrub through historical weeks

Simulation answers a specific question. Motif World provides the broader evolving backdrop that scenario work draws on.

Last updated