This directory preserves active laboratories, exploratory prototypes, and historical development lineages of NEXAH.
It is a workspace for investigation, not a second operational core. Content in
this directory may be incomplete, duplicated, speculative, or superseded by
material in RESEARCH/, PROTO_CORE/, ARCHITECTURE/, or APPLICATIONS/.
Each area is assigned one of four working statuses:
- Active Lab — a current, bounded experimental investigation
- Experimental System — runnable or partially runnable prototypes that require review before reuse
- Historical Reference — conceptually useful earlier architecture or theory
- Archive — preserved development history; not a current entry point
Experimental results are not automatically validated findings. Promotion into another repository layer requires a clear claim, reproducible evidence, and an explicit destination.
| Area | What it contains | Status | Recommended use |
|---|---|---|---|
| OBSERVER_GEOMETRY_LAB/ | Observer-relative geometry, projections, transport structures, and visual experiments | Active Lab | Best current experimental entry point |
| BUILDER_LAB/ | Demos, simulation engines, discovery work, exploration notes, and several archived systems | Experimental System / Archive mix | Enter through the guided map below |
| FRAMEWORK/ | Earlier META → ARCHY → MESO → NEXAH → MEVA conceptual architecture | Historical Reference | Consult for conceptual lineage, not current implementation state |
| Legacy Root Outputs | Frozen generated figures and result files formerly exposed at repository root | Archive | Historical inspection only; not current evidence or capability status |
The empty root scripts/ and visuals/ directories are not current entry
points.
OBSERVER_GEOMETRY_LAB/ is the most clearly bounded laboratory in this directory. It investigates:
- local and global representations
- observer-relative projection
- transport corridors and reinjection structures
- manifold slicing and orientation axes
- Mandelbrot–Julia contact visualizations
The lab includes its own README, documentation, scripts, and selected visuals. Its claims remain exploratory and have not been promoted into the validated Research layer.
BUILDER_LAB/ contains most of the directory's material and should not be read
as one coherent runtime. It combines several development generations.
| Builder Lab area | Role | Current interpretation |
|---|---|---|
| demos/ | Small state-graph and navigation demonstrations | Test before presenting as user tools |
| engines/ | Synthetic infrastructure, cascade, planetary, and multi-system simulators | Application prototypes, not validated applications |
| systems/ | Example system definitions | Inputs for Builder Lab experiments |
| dashboards/ and visualizers/ | Experimental interfaces and renderers | Prototype tooling |
| DISCOVERY_ENGINE/ | Early transition, resilience, field, and law-discovery experiments | Historical experimental lineage with promotion candidates |
| EXPLORATION/ | Control notes, symbolic work, fragments, older portals, and experimental scripts | Mixed research notebook; review item by item |
| proto_models/ | Oval-membrane and time-knot conceptual models | Early concept models |
| ARCHIVE_KERNEL/ | Earlier field, navigation, and kernel variants | Archive |
| ARCHIVE_ENGINE/ | Large historical engine, research, navigation, and output tree | Frozen development archive |
The existing Builder Lab Inventory is a current structural audit dated July 12, 2026. It records entry-point checks, dependency boundaries, archive weight, and promotion candidates. It is an inventory rather than a software specification.
The archive contains several earlier attempts to turn the NEXAH idea into a kernel or application system. They are useful for understanding how the current architecture emerged. Their presence does not make them part of the maintained runtime.
| Lineage | Start here | What is actually present | Current relevance |
|---|---|---|---|
| Early kernel shell | Archived Kernel README and NexahKernel | A layered graph → regime landscape → navigation → action architecture; parts of the navigation implementation remain placeholders | Preserves the early system decomposition |
| Discrete Navigation Kernel v2 | Navigation README and navigator.py | Executable graph lookahead, risk-distance scoring, and next-state selection using supplied regimes and risk targets | Useful algorithmic reference; labels and scores require independent evidence before reuse |
| Field and Dynamics Engine | Dynamics Engine README | Flow, basin, topology, transition, phase-map, and meta-field experiments across many development levels | Hypothesis and method archive; not the current orientation backend |
| Stability-driven multi-agent study | Experiment README | A multi-agent research concept and visual centered on stability-seeking exploration | Reconstruct only for a specific testable question; current evidence is insufficient for a finding |
| Prototype agent runner | agent_run_demo.py and NexahAgent | Grid-landscape exploration, a reward-based learning prototype, and a separate skeletal agent loop | Reference for experimental orchestration, not an integrated agent layer |
| Discovery Engine | Discovery Engine README and Discovery Observations | Resilience, transition, field, phase, and symbolic-law experiments | Selected observations may be reproduced; universal-law language is not a validated capability |
| META–ARCHY–MESO framework | Historical Framework | Earlier conceptual layers connecting purpose, structure, dynamics, navigation, and action | Conceptual lineage only; the current contracts use a different architecture |
Two related reference areas now live outside EXPERIMENTAL/:
- Dynamical model hierarchy documents the conceptual progression from stability landscapes through gradient and drift dynamics to regime systems.
- Control Layer preserves executable synthetic gate, field, routing, and multi-agent prototypes. Its own status explicitly limits them to experimental trajectory-deformation and geometry-aware routing studies.
Across these generations, one architecture repeatedly reappears:
system observations
→ representation
→ structure or landscape
→ states and transitions
→ orientation and navigation
→ optional intervention
The maintained Orientation Layer keeps this structural direction while adding typed contracts, provenance, uncertainty, reproducibility, evidence gates, and a strict boundary between orientation and execution. Historical components should therefore be treated as design history or candidates for bounded reconstruction, not imported as a second kernel.
The following entry points exist, but still require a dedicated runtime and dependency check:
python EXPERIMENTAL/BUILDER_LAB/run_builder_lab.py
python EXPERIMENTAL/BUILDER_LAB/nexah_cli.py systems-list
python EXPERIMENTAL/BUILDER_LAB/demos/nexah_demo.pyFor the verified repository quickstart, use the NEXAH Demonstrator instead.
FRAMEWORK/ documents an earlier layered model:
META → ARCHY → MESO → NEXAH → MEVA
It provides conceptual background for coherence, risk fields, regimes, transition geometry, and navigation. These layers are not the current software architecture. For the implemented repository view, see the Architecture Index and System State.
The following items appear worth reviewing outside the archive. Inclusion here means candidate for assessment, not accepted result.
| Candidate | Possible destination | Required review |
|---|---|---|
| Discovery Observations | Remain experimental | Reviewed July 12, 2026; V22 needs controlled reproduction before Findings promotion |
| Symmetry Graph result summary | RESEARCH/FINDINGS/ |
Verify code, data, metrics, and current terminology |
| Structured Oscillator Networks | RESEARCH/FINDINGS/ or RESEARCH/CONCEPTS/ |
Determine whether it contains evidence or only a model proposal |
| Stability-driven Multi-Agent System | RESEARCH/FINDINGS/ |
Reproduce the experiment and qualify the stability claim |
| Control Sensitivity Field | RESEARCH/CONCEPTS/ or ARCHITECTURE/METHODS.md |
Formalize definitions and establish relationship to current control prototypes |
| IEEE Mapping Module | Remain historical | Reviewed July 12, 2026; conceptual analogies are superseded by current power-system experiments |
| Oval Membrane Field | RESEARCH/CONCEPTS/ |
Decide whether the model yields testable hypotheses |
| Time Knot Field | RESEARCH/CONCEPTS/ |
Separate mathematical proposal from metaphorical interpretation |
| Builder Lab demos | APPLICATIONS/ index |
Test dependencies, outputs, and user value before exposing them as tools |
| Observer Geometry Lab | Remain here; later link from Research | Define validation questions before any promotion |
Promotion should normally copy or rewrite the smallest relevant unit and retain a source link. Moving an entire historical tree would erase useful context and carry obsolete assumptions into current documentation.
Before experimental material becomes a current finding, method, or application, check that it has:
- a precise question or claim
- a runnable or otherwise inspectable method
- identifiable inputs and generated outputs
- limitations and failed cases
- terminology consistent with the current Gate and transition interpretation
- a clear owner directory: Research, Proto Core, Architecture, or Applications
If these conditions are not met, the material should remain here and be indexed as an experiment or historical reference.
EXPERIMENTAL
→ generates prototypes, hypotheses, and exploratory evidence
RESEARCH
→ frames questions, records findings, and organizes validation
PROTO_CORE
→ develops and exposes reference implementations
ARCHITECTURE
→ explains system relationships and implementation maturity
APPLICATIONS
→ presents system-specific tools and studies
Related entry points:
NEXAH Experimental
Active Labs · Prototype Systems · Historical Lineages · Promotion Candidates