What it is
SignalChain is a Python framework for data cleaning and categorical variable analysis. It recognizes field types (gender, age, amount, etc.) in dirty CSVs and classifies categorical variables, using AI for decisions and local scripts for execution.
How it uses Jev
Jev (System One) is an optional fast path for the cleaning pipeline. It answers Choice questions to classify each CSV column into a 13-code type vocabulary and each dataset into one of six scenes. Results are used as signal_sequence to drive local cleaning operations. Low-confidence fields default to X (pass-through).
Primitives:choice
Technique worth stealing
Use a typed Choice model to output one character per column, avoiding token waste from repeated criteria.
Try it
uv sync --extra system1 && uv run python run_clean.py medical --system1
Evidence
Each line is one question put to Jev about the README. ≥ 0.60 reads as yes, ≤ 0.40 as no; in between Jev is not making a call.
- Jev-centricno0.16
- Shows a System One patternyes0.76
- Handles uncertaintyno0.04
- Measuredyes0.65
- Runnableyes0.89
- Worth recommendingno0.39
- Model replicano0.08
- Problem scopescore on a 0–2 scale1.66
- About Jevyes0.96
Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 21 Sept 2026.