Skip to content
Metatext Decide — typed decisions for your agents, with a probability on every answerGet early access
Metatext
Sign Up

jev-riffs

MIDI motif ripper: code mines repeating interval patterns and one batched Jev call grades each candidate's musical significance.

ProjectprototypeMIT
View on GitHub ↗

Overview

jev-riffs is a music pattern ripper. It tokenises a MIDI file into a string of semitone intervals, mines it in code for repeating units, and has Jev grade each candidate's musical significance in one batched request. A web UI draws a piano roll, highlights where each motif occurs, plays the passage on click and exports WAV or JSON.

The interval encoding is transposition-invariant, so a riff restated in another key produces the same token substring. Candidate mining is pure code: it counts non-overlapping repeats and prunes shadows, periodic units, rotation echoes and superstrings that do not add coverage, ranking by coverage. The README reports that on a 17-track orchestral MIDI of Swan Lake, the ripper returned the main theme as its top-ranked motif in one batched call in about 600 ms. A reusable Jev client library is documented separately.

What the decision model decides

TypeQuestionAnswers
ScoreHow musically significant is this candidate motif?—
Yes / noIs this candidate a meaningful pattern?yes · no

How it works

  1. Parse MIDI note-ons and pick the melody track
  2. Convert to interval tokens
  3. Mine and prune repeating candidates in code
  4. Grade candidates with one batched Jev request
  5. Show results in a piano-roll UI

Strengths

  • Deterministic mining with model only for judgment
  • Transposition-invariant pattern detection
  • Fast single-request grading

Caveats

  • Melody heuristic picks one track
  • Significance judgment is subjective
  • Requires a TypeSafe key

At a glance

Type
Project
Status
prototype
Language
TypeScript
Runtime
Web app
Has code
yes
License
MIT

Use cases and integrations

How curated lists describe it

  • Music pattern ripper: MIDI → interval tokens → code mines candidate motifs → Jev (TypeSafe System One) grades their significance. Web UI with piano roll, click-to-play, WAV export.

    logicrw/awesome-jev-projects

  • Music pattern ripper: MIDI → interval tokens → code mines candidate motifs → Jev (TypeSafe System One) grades their significance. Web UI with piano roll, click-to-play, WAV export.

    hellogumbo/awesome-jev

Frequently asked questions

What does Jev decide?

How musically significant each code-mined candidate motif is.

Does Jev find the patterns?

No. Code mines repeating patterns; Jev only grades them.

Does it work on real music?

The README reports extracting the Swan Lake theme from a 17-track MIDI.

Related projects

  • jevclip

    CLI that has Jev judge a video's subtitles segment by segment, then cuts a highlight reel and a filler-free full version.

    ProjectbetaPython
  • ldraw-nova

    Agentic LEGO builder that turns an idea into an LDraw model, using Jev reranking to find suitable parts and example models.

    ProjectbetaPython
  • ReelQL

    Claude skill and API that turns a video link into timestamped JSON, then asks Jev typed questions to rank or filter videos against a brief.

    Agent skillbetaPython
  • jev-creator-decoder

    Cloudflare app that labels a YouTube creator's Shorts with Laya or Jev to find which hooks drive views and score new hooks.

    ProjectprototypeTypeScript
  • jev-in-blender-experiment

    Blender extension where Jev ranks about 2,500 operators against a plain-language request via a two-stage Choice search.

    ProjectprototypePython
  • JEVMETER

    CLI that transcribes a video, has Jev score every sentence against rubric flags, and renders a 16:9 edit with a live meter.

    AppbetaPython