SignalProof

An evidence-first low-voltage signal quality workbench that computes six deterministic metrics (DC offset, AC RMS, rail clipping, 48-52 Hz mains hum, timing jitter and capture duration) before validating bilingual explanations against fixed finding IDs.

Period
November 2026
Release
Public source project
Work type
Independent open-source project
Development
AI-assisted, with verification
SignalProof project preview
Project preview restored from its existing public asset.
01 / Overview

What I built

An evidence-first low-voltage signal quality workbench that computes six deterministic metrics (DC offset, AC RMS, rail clipping, 48-52 Hz mains hum, timing jitter and capture duration) before validating bilingual explanations against fixed finding IDs.

Developed with AI coding assistance. Source code, reproduction instructions, a 3-minute narrated walkthrough on YouTube and tests are available via the links above.

02 / Evidence

What was checked

The repository includes deterministic Node.js checks and synthetic evaluation fixtures. Its archived telemetry manifest records 520 synthetic inference calls and 1,313,133 tokens. These are recorded experiments, not new model calls or production accuracy measurements.

SignalProof workflow diagram
Workflow illustration, not a screenshot of an executed run.

Verification date: 7 November 2026. Tests cover the included fixtures and recorded audit traces and do not establish production performance.

03 / Scope

What this demonstrates

Validates signal quality rules on synthetic and CSV bench captures (0.0-3.3 V default range). Passing deterministic checks and recorded LLM audit replays do not establish calibrated oscilloscope accuracy or hardware safety.

All projects

Source audit

Public source reviewed at the linked commit. Tests describe their included fixtures and do not establish field or production performance.

15 November 2026. Reviewed source 0926e62a39c9. Source evidence index.

Navigate with arrow keys. Enter opens a result. Escape closes search.