Learn how

Trust, built in seven layers

A unified framework for trust-based decision making — from a single seller to a committee of agents. Learn the model, then look up the API.

Core T_eff = τ · P · Φ(δ)

Dyadic & collective Bayesian learning Game-theoretic override Zero dependencies

Documentation

Choose your path

Two documents, one framework. Start with the tutorial if you are new; jump to the guide if you need a specific answer.

Read the Tutorial

Narrative Introduction

A narrative introduction that builds the framework from first principles — no mathematics prerequisites.

Why trust needs seven layers, not one number
Each layer explained with intuition, formalism, and code
The Group Trust Extension and the Unification Theorem
Seven hands-on application walkthroughs

Open tutorial

~30 min read 7 walkthroughs

Browse the User Guide

Task-Oriented Reference

A task-oriented reference to every class, method, parameter, and recipe in trustlib_mas.

Full API reference for the four core classes
Parameter calibration and troubleshooting
Common recipes and advanced patterns
Simulation, ablation, and robustness toolkit

Open user guide

Reference 24 chapters

Seven composable layers

Use the ones you need. Each is independent, and together they form a complete decision pipeline.

L1Perceived Trustworthiness L2Decision Threshold L3Bayesian Learning L4Temporal Dynamics L5Strategic Override L6Network Propagation L7Group Trust Extension