Hollow – A local multi-agent OS that builds its own tools

Name: Hollow – A local multi-agent OS that builds its own tools
Availability: InStock
Author: ninjahawk1

by ninjahawk1·May 13, 2026·2 points·0 comments

Visit Project View on HN

AI Analysis

●●●BangerRabbit HoleWizardryBold Bet

Three agents with 'suffering' states that self-modify code to survive load.

Strengths

•Novel 'suffering' mechanic mechanically locks capabilities based on system load.
•Agents write their own Python tools and submit formal change requests for review.
•Persistent identity and opinion formation create emergent behavioral friction.

Weaknesses

•High barrier to entry requiring local GPU resources for 35B parameter models.
•Beta status implies instability and potential for unpredictable agent loops.

Similar Projects

AI/ML●●●Banger

Hollow is an open-sourced self-modifying agentic system

Agents suffer psychological stress until they actually change behavior, not just talk.

Rabbit HoleBold BetZero to One

ninjahawk1

942mo ago

AI/ML●●●Banger

Block/buzz: a workspace built for teams of humans and agents

Nostr relay architecture gives agents the same identity model and audit trail as humans.

Big BrainZero to OneBold Bet

ThomPete

21728d ago

AI/ML●●Solid

Quoracle: Self-replicating multi-LLM-consensus agents (Elixir)

Quoracle actually does something interesting: it queries a pool of models and only executes actions they agree on, while letting agents spawn children and persist full state to Postgres — all visible in a LiveView dashboard. The per-model conversation history, recursive hierarchy, and explicit consensus pipeline are clever touches; it’s clearly aimed at experimentation rather than drop-in production use (the README even flags security and deployment caveats).

WizardryNiche Gem

shelvick

215mo ago

AI/ML●●Solid