Deterministic floor, model ceiling

An agentic OS for
LLM-centered development.

Zunstack runs a fleet of projects — research, games, fiction, benchmarks — under one harness that gives every AI session memory, gates, and bookkeeping it can't skip.

the fleet — live topology · click a node

Doctrine

Deterministic floor, model ceiling.

The model decides; deterministic tooling enforces and records. Nothing that records or projects state ever calls a model — records are pure functions of the op log, so every projection reproduces.

01

Zunstack / engine

An agentic OS: harness crafting, memory, gates, bookkeeping.

active dev · 3 monthslast updated · 2026-07-18
moduleprovenancecadencecarried by
structures-indexR — regeneratedimmediateall (universal map)
harness-structureP — authoredbatchall
oplog / op-reportshybrid P+Rimmediateall
behavior · convergenceP — model-synthesizedbatchwiki-type
standing · docs-indexP — authoredbatchwiki-type
code_moduleR — code-cyclebatchcode repos
trendlineR — op-series metricsbatchplanned
contractP — tracked contractown cycleall

Sessions forget. The fleet doesn't.

02

Claude-control

The Zunstack app — 99% of user interaction lives here.

active dev · 2 monthslast updated · 2026-07-18

the plane — real capture

One surface for every session — desktop, phone, village.

03

LLM-research

A database that compounds — not a pile of search results.

active dev · 1 monthlast updated · 2026-07-08
layercontentsstate
rawimmutable captured sourcesbuilt
source pages + derivativesdistilled summaries · claim-level provenance (31 files)built
concepts / tools~76 interlinked user-facing pagesbuilt · growing
synthesis + reportsuser-led insight · daily/weekly movement reportsnext

Knowledge that compounds while you sleep.

04

Self-writing LLM

Event sourcing applied to fiction.

active dev · 1 monthlast updated · 2026-06-12

A story world you can revert like code.

05

ARPG sandbox

A game development suite with a playable co-op ARPG inside.

active dev · 1 monthlast updated · 2026-07-18

live from the workshop — real capture

in-game — real capture

[ dungeon forge — scroll to load ]

the real generator + renderer, running live

Build system, identity system, and endgame share one geometry.

06

Bench / ML

A personal measurement instrument for model quality.

last updated · 2026-07-15

3,111

rows scored

2,882

loop points

405

failures caught

$40.11

total cost

suitesignalscoring
IFEval + IFBenchinstruction adherencedeterministic checks
BBH (hard subset)reasoning under thinkingexact match
GPQA-Diamond + HLEraw capabilityexact match
ARC-AGI-2abstract reasoninggrid equality
Aider-Polyglotagentic codingtest suite pass
comms_recall (custom)buried-detail survivalrecall scoring — opus 0.85 · haiku 0.53

Vibes are not data. This is data.

07

Endgame

The projects aren't a portfolio. They're organs of one machine.

End-state Zunstack is an operating system for intent: you say what you want, and the fleet does the rest. Sessions spawn under harnesses built for their work. Choices get made from taste profiles — your decision-making, as data — instead of waiting on you. Quality is measured, never vibed. Prompts are recursively refined against bench scores. Knowledge compounds from the web while you sleep. And every layer feeds the next: the research informs the refinement, the refinement sharpens the sessions, the sessions build the tools that tighten the loop.

engine

rigor for semantic work

Code has compilers and tests; prompts, contracts, and instructions have nothing. The engine gives non-mathematical work the same discipline — gates, records, validation loops.

bench

testing harnesses, prompts, variance

You can't improve what you can't measure. Bench turns harness changes, prompt edits, and model variance into scored, confound-stamped data.

ML

recursive improvement

Bench deltas become training signal: prompts recursively refined against real scores instead of intuition. The optimization loop over the whole stack.

LLM-research

fresh knowledge, compounding

Models go stale and web searches don't accumulate. A vetted, revertible database that scans in loops and compounds — feeding the improvement cycle fresh insight 24/7.

claude-control

the surface + automated choices

A human approving every step is the bottleneck. Taste profiles — decision data derived from engineers — let runs make your choices without you in the room.

worldbuilding vault

storybuilding as a system

AI fiction fails on continuity — docs full of info don't stop contradictions. Event-sourced, delta-gated, reversible world state makes long-form story a solved data problem.

ARPG sandbox

the code-native proving ground

The other pole of the spectrum: where compiler + tests + golden corpus already make drift loud, the harness sheds machinery — proving the module system adapts to its feedback signal.