Skip to content

Jev directory

Jev Music Tag

Minimal Jev metadata workbench for local music files: a FastAPI backend sends track metadata and scoring questions to TypeSafe Jev and can write the chosen genre or language back with Mutagen, behind a small React page for editing prompts and questions.
Stars
6
Language
Python
Reading time
3 min
Published by
Community
Author
xhongc
Last verified
Oct 2026

Highlights

  • FastAPI POST /api/decide sends track metadata plus scoring questions to TypeSafe Jev's systemone endpoint and returns metadata_updates ready to write back.
  • Mutagen reads and writes tags on MP3, FLAC and M4A files on the server; separate POST /api/read and POST /api/write endpoints handle tags directly.
  • Empty values are stripped and album_img excluded before anything reaches the model; only allowlisted writable metadata fields can be a target_field.
  • The React page is a capsule-style workbench where you edit the JSON, prompt and question configuration; the server picks the top probability or score.
  • Config lives in .env: JEV_API_KEY is required and server-side, JEV_BASE_URL defaults to api.typesafe.ai and JEV_MODEL_NAME to jev-latest.

Quickstart

bash
cp .env.example .env
uv sync
uv run uvicorn backend.main:app --reload --port 8000

Watch out

Copy reviewed against the Chinese README on 2026-10-15; an English README.en.md is linked. The README itself calls this a minimal demo: no database, login or batch scanning, and production use would need auth, a path allowlist, log redaction and rate limiting. GitHub metadata lists no license, and typed file paths refer to the server machine.

Questions about this entry

Short answers drawn from the linked project or page.

From the Jev guides

Original write-ups on what this entry is for.

From the Jev AI model community

Posts from builders shipping with Jev right now.

Follow @typesafeai