How it's wired Tepna
Tepna · the -Dex suite · internals

How it's wired

🗺️ Architecture — start here 📄 CONTRIBUTING.md

One reference per receptor. Each page explains what the node senses, the file stack that powers it, the data flow from file-drop to Ganglior event, how far to trust each metric, the honest limitations, and how to use it — written to be read by humans and AI alike.

receptors the -Dex nodes Ganglior the relay bus Integrator fusion insight the read-out
THE NODES

Seven receptors, one signal each

A -Dex node is an afferent receptor: it senses exactly one physiological signal and emits events onto the Ganglior bus. Nodes never import each other — they meet only through the export contract, fused by the Integrator. Pick a node to see its wiring.

RESEARCH & VALIDATION

The tier that proves the suite

Beneath the instruments sits the research scaffolding: a synthetic-corpus generator with known ground truth, a harness that runs the real detectors at scale, and the preprint series that writes up what they find. Three more wiring guides cover it.

READING THESE DOCS

What every page covers

What it's for§01

The one signal the node senses, its device, input format, and its role as an afferent receptor in the reflex arc.

The wiring§02

The external *.js file stack — DSP, Core, Render, App, Fusion — each card a file, each with its one job and the downhill dependency direction.

How it's achieved§03

The pipeline from file-drop → clock-stamped parse → DSP compute → graded render → Ganglior emit, with the real function names.

Evidence + limits§04–05

Each metric on the five-tier evidence ladder, then the honest, node-specific limitations — what the signal can and can’t say.

These are reference docs, not bundled instruments — they load no app code and carry no buildHash. The apps themselves are frozen, hash-verifiable single files; edit the .js + .src.html and re-bundle to change one.