Skip to main content
A simple and universal swarm intelligence engine

Upload Any Report. Simulate The Future Instantly.

MiroFish transforms real-world seeds such as news, policy drafts, analytical reports, or fiction into a parallel world populated by agents with memory, motives, and social behavior.

From public opinion forecasting to literary continuation, it lets you inject variables, run social evolution, and inspect the likely trajectory before the real world catches up.

Watch Demo
1M+
Parallel agents per run
Twitter + Reddit
Simulation surfaces
Prediction report
Primary output
Hero Surface
MiroFish
v0.1 preview
MiroFish visual
Workflow Sequence
Step 01

Ontology Generation

Turn raw reports, notes, or fiction into structured entities, motives, and factual anchors.

Step 02

Graph Construction

Assemble a living relationship graph that exposes the actors, tensions, and memory structure behind the scenario.

Step 03

Parallel Simulation

Let platform-native agents interact across Twitter and Reddit style channels over multiple rounds.

Step 04

Report Generation

Condense the trajectory into a readable prediction report with key turning points, risks, and confidence signals.

Step 05

Deep Interaction

Interrogate the generated world through ReportAgent or by interviewing individual characters directly.

System Status

Engine Standing By

Upload narrative seeds, analytical reports, policy drafts, or story fragments. The graph, simulation, report, and interaction chain will be opened from one console.

Sequence
01
Ontology Generation
Turn raw reports, notes, or fiction into structured entities, motives, and factual anchors.
02
Graph Construction
Assemble a living relationship graph that exposes the actors, tensions, and memory structure behind the scenario.
03
Parallel Simulation
Let platform-native agents interact across Twitter and Reddit style channels over multiple rounds.
04
Report Generation
Condense the trajectory into a readable prediction report with key turning points, risks, and confidence signals.
05
Deep Interaction
Interrogate the generated world through ReportAgent or by interviewing individual characters directly.
Engine Console
Seed the simulation
PDF / MD / TXT
01 / Reality Seeds
02 / Simulation Prompt
History Database

Simulation history

FAQ

Questions before you simulate

What can I upload into MiroFish?

The current product shell accepts PDF, Markdown, and plain text. The workflow assumes the file contains enough narrative, analytical, or factual seed material to build a graph and simulation context.

Does this really simulate multiple platforms?

The product flow is designed around a dual-surface simulation. In fixture mode the posts and actions are deterministic; in backend mode the Python service is responsible for generating platform-native behavior.

Can I use it for public opinion and fiction?

Yes. The built-in scenarios cover public opinion events, literary continuation, market or finance cases, and a general-purpose predictive mode.

What does the report include?

Each report includes an executive summary, predicted developments, major risks, evidence lines, key actors, and a long-form narrative explanation of how the outcome unfolds.