PeerWeight Protocol

A Decentralized Trust and Discovery Protocol for the Human Web

Overview

PeerWeight Protocol creates a decentralized trust mesh through cryptographically signed endorsements. Unlike traditional search engines that optimize for engagement, PeerWeight enables trust-based discovery where each user maintains their own personalized View—a trust graph that returns different results based on who they trust.

Key Features

• Decentralized Trust: No canonical ranking, each user has their own personalized view
• Cryptographic Integrity: All endorsements are cryptographically signed
• Interoperable: Standards-grade protocol suitable for independent implementations
• Privacy-Focused: Users control their own data and trust relationships

Protocol Primitives

The protocol defines four core primitives:

• Identity: Who you are (DID-based identities)
• Endorsement: What you recommend
• Note: What you think (threaded discussions)
• View: Who you trust (personalized trust graph)

Documentation

Core Documentation

• PeerWeight Protocol Specification v1.3.6 - Full protocol specification
• Whitepaper - Research paper on implementation, trust propagation algorithms, Sybil resistance, and AI agent integration (~15,000 words)
• Implementer Notes - Implementation guidance and best practices

Developer Resources

• AI Agent Quick Start - Build a PeerWeight client in 10 minutes
• JSON Schemas - Machine-readable schemas for all primitives
• OpenAPI Spec - REST API specification for aggregators
• llms.txt - LLM-optimized protocol summary

Getting Started

Installation

bun install

Running the Aggregator

cd peerweight-aggregator
bun run src/index.ts

Project Structure

• PeerWeight-Protocol-v1.3.6.md - Protocol specification
• IMPLEMENTER_NOTES.md - Implementation guidance
• peerweight-aggregator/ - Reference implementation of an aggregator service

Links

• Website: peerweight.org
• Author: Derivative Labs

License

Version 1.3.6 - January 2026