Nova

Philosophy

Engineering stability should be assessed separately from scientific effectiveness. Reproducible stage progress and content provenance are necessary, while real models, retrieval and research experiments are still needed to assess value.

Architecture & control flow

1. 01

   Before the first candidate exists, keep mathematics, physics and mapping branches from dominating the main path.

2. 02

   Control the mathematics–physics bridge with deterministic depth conditions and branch quotas.

3. 03

   If experiment or attack nodes exist without a candidate, make a limited number of terminal-progress steps.

4. 04

   Check global-node and review budgets during repair and distinguish seed drafts from candidates with sufficient model fields.

Architecture outline derived from this version’s control flow.

What this version changes

Compared with Cosmos, addresses random triggering, candidate-path starvation and repair-budget growth, with separate checks for seed drafts, model-filled candidates and seed-only runs.

Implementation & evidence scope

The changelog reports simulated workflow-stability tests, not stable production of scientifically novel results. Bridge quotas relax after a candidate exists and terminal progress can exceed the main-loop node count, so this is not an absolute budget bound. Real eight-hour model runs and full live literature retrieval remain unverified.

Code & bundled material

The introduction draws on bundled notes, changelogs and central code. Software tests, synthetic diagnostics and scientific effectiveness use different evidence standards.