Logos

Philosophy

Evidence must refer to the same candidate before its quality can be assessed. Broken package links are classified separately from scientific refutation.

Architecture & control flow

1. 01

   ArtifactIndex joins AST, simulation, quality, proof, review, delta and path-trace records by mapping ID rather than positional zip.

2. 02

   Real D/M/P traces drive compatibility and prior scoring; PropertyProvenance labels default, path, mapping, simulation and proof-derived properties.

3. 03

   Matched elite evidence creates DiscoveryIdeaV58 and an explicit artifact graph.

4. 04

   Aurora-V58 adds G-CHAIN to seven evidence gates; broken links produce PACKAGE_INVALID instead of scientific refutation.

5. 05

   Dictionary/dataclass replay equivalence and one authoritative top-level package support consistent serialization, with older blocks retained under legacy.

Architecture outline derived from this version’s control flow.

Inputs & outputs

Inputs

Genesis elites, actual traces, ID-matched equation/simulation/quality/proof/review artifacts, property provenance and replay parameters.

Outputs

One discovery_package.json, candidate-specific ideas, artifact graph/index, Aurora-V58 statuses, package audit/metrics and replay-equivalence checks.

Implemented components

• ID-based candidate evidence chains, property provenance/caps, real path-trace adapters, PACKAGE_INVALID classification and serialized-AST replay support.

Implementation & evidence scope

A real LLM D-Tree does not yet drive the main search; the retained proposer is deterministic, while richer ASTs and ontology evolution remain later milestones. PROMISING_BUT_UNPROVEN is a protocol status, not externally verified originality, machine proof or literature validation. The release adds no solver, mapping family or conjecture type and should not be described as additional scientific discoveries.

Code & bundled material

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