jevbooks

← All projects

sqlite-jev

mgaitan/sqlite-jev

Batched natural-language judgments for SQLite via TypeSafe Jev.

Classification & Routing32%Runner-up: Scoring & Ranking

What it is

A SQLite loadable C extension plus Python package that brings TypeSafe Jev to SQLite. It exposes Noul, Choice, and Score judgments as SQL scalar functions and batched virtual-table queries, aimed at analysts and developers doing natural-language classification or scoring over table data.

How it uses Jev

Jev makes Boolean Noul judgments (probability of yes), Choice classification, and Score ratings. State is a JSON object built from selected table columns; up to 40 rows share one API state. Results return as virtual-table rows (answer, probability, choice, score, confidence) joinable by source_rowid, or via scalar functions.

Primitives:noulchoicescore

Technique worth stealing

Batch up to 40 rows in one shared Jev state via jev_rows, keeping thresholds in SQL so they can rise with false-positive cost.

Try it

Run make, export TYPESAFE_API_KEY=..., then .load ./build/jev in sqlite3 and call jev_version().

View on GitHub

judged by Jevjev-1.13.0

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-centricyes0.95
  • Shows a System One patternyes0.86
  • Handles uncertaintyno0.19
  • Measuredno0.05
  • Runnableyes0.97
  • Worth recommendingunclear0.50
  • Model replicano0.12
  • Problem scopescore on a 0–2 scale1.13
  • About Jevyes0.98

Signals by Jev jev-1.13.0, card written by DeepSeek V4.1 Flash from the README on 20 Sept 2026.