systems-explained / hash-labs

🧪 Public laboratory for experiments

Labs

A public laboratory for HASH's experiments 🧪

This repository contains code snippets, examples and experimental work made public. Projects from our private internal-labs (🔒) monorepo for experiments will be moved here and published when safe and practical to do so, in line with our commitment to open-source.

Apps

The apps directory contains the source-code and/or content for a number of runnable applications.

Simulation Tools

The HASH app seeks to enable its users to make better decisions by utilizing all of the information available to them. Generative simulation is a core part of realizing this vision. In anticipation of integrating these capabilities, we have developed a variety of standalone tools for agent-based modeling. These simulation tools remain experimental, but we are open to accepting contributions to these in line with our repository guidelines.

• sim-core is an open-source, locally-runnable version of HASH Core
• sim-core-plugins contains an example external plugin developed for hCore, which provides a visual interface for process modeling
• sim-engine contains HASH Engine, a versatile agent-based simulation engine written in Rust (with support for TypeScript & Python sims)

POCs

The pocs folder contains proof of concepts and other one-off experiments.

• distributed_collab - Distributed Collab: a BEAM VM based system for publishing/subscribing to JSON Patches with core logic implemented in Rust
• hash-agents - HASH Agents: an experimental setup for writing Python-based 'agents' that interface with LLMs
• hash_helm_chart - HASH Helm Charts: An experimental Helm chart for deploying an instance of HASH on Kubernetes

A number of older POCs can be found in our hasharchives organization, including:

• wasm-ts-esm-in-node-jest-and-nextjs - A Wasm + TypeScript + ESM in Node.js, Jest and Next.js 13 example project