When should I use the tropical distance metric?

Use it when you need to determine the most efficient sequence of skills to execute a task while maintaining mandatory redistribution of information between agents.

What role does GF(3) play in this skill?

GF(3) accounting treats information as trits where validators are sources (-1), coordinators are channels (0), and generators are sinks (+1), ensuring the system maintains mathematical balance.

What is a fixed-point pathology in AI workflows?

A fixed-point pathology occurs when an agent or validator validates its own output or patterns, creating a feedback loop that fails to catch errors or improve information quality.

How does the derangement constraint improve code reviews?

By enforcing that Agent A validates Agent B's output rather than its own, the derangement constraint ensures that every bit of information must flow to a different position, increasing oversight integrity.

Derangement Reflow Control

Name: Derangement Reflow Control
Author: plurigrid

byplurigrid

•

开发者工具

Optimizes agentic workflows by enforcing derangement constraints to prevent circular validation and ensure healthy information flow.

Derangement Reflow implements a mathematical framework for multi-agent coordination based on derangement theory and tropical geometry. It is designed to eliminate 'fixed-point pathologies' where agents inadvertently validate their own logic or patterns. By applying GF(3) triadic accounting, the skill ensures information moves dynamically between Validators (entropy sources), Coordinators (channels), and Generators (entropy sinks). This approach leverages active inference and joint world modeling to minimize free energy across complex tasks, making it ideal for high-integrity software development and complex automated reasoning.

主要功能

01Calculates optimal skill interleaving using tropical geometry metrics

02Identifies and fixes fixed-point pathologies in PR reviews and task routing

03Enforces derangement constraints to prevent circular logic in agent workflows

048 GitHub stars

05Implements GF(3) triadic accounting for information flow conservation

06Facilitates joint world modeling via active inference principles

使用场景

01Detecting self-validation biases in automated code review and quality assurance processes

02Optimizing complex task interleaving to minimize the 'tropical distance' between operations

03Designing multi-agent pipelines where independent validation is mathematically required

What are Skills?·How to Install

Install with 🐟 Skill.Fish

npx skillfish add plurigrid/asi derangement-reflow

For use in Claude.ai and ChatGPT

Download Skill