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

patrol-jev

Korean patrol-log tool: upload photos, a general AI reads them and Jev sorts each note into one of four categories for the log.

DemobetaMIT
View on GitHub ↗Live demo ↗

Overview

patrol-jev turns patrol photos into patrol-log text and Hancom .hwpx files for Korean local administration, and compiles a month of logs into a results report of counts and locations without rankings or evaluations. The README is in Korean. It groups photos by location, reads address plates, and assembles sentences automatically; points that need judgment are left for the staff member.

Two keys are used: OpenAI transcribes photos into text, and Jev sorts that text into one of four categories; grouping and sentences are done by code. A manual mode calls no model at all, keeps photos in the browser and costs nothing. There are no accounts or databases, records stay on the local computer, and it is MIT licensed. A hosted test page has a daily limit.

What the decision model decides

TypeQuestionAnswers
ChoiceWhich of four categories does this patrol note belong to?—

How it works

  1. Upload patrol photos.
  2. OpenAI reads each photo into text.
  3. Jev assigns one of four categories.
  4. Code groups by location and writes the log and .hwpx file.

Strengths

  • Manual mode works without keys or cost.
  • No accounts or database; local records.
  • Saves routine writing for staff.

Caveats

  • Korean-specific workflow and documentation.
  • Model mode needs both OpenAI and TypeSafe keys.

Getting started

git clone https://github.com/patrol-jev/patrol-jev.git && cd patrol-jev && npm install && cp .env.example .env.local && npm run dev

At a glance

Type
Demo
Status
beta
Language
TypeScript
Runtime
Next.js web app
Has code
yes
License
MIT

Use cases and integrations

How curated lists describe it

  • 사진을 올리면 순찰일지 글이 나옵니다. 읽기는 범용AI, 판단은 Jev 가.

    hellogumbo/awesome-jev

Frequently asked questions

What does Jev decide?

Which of four categories each patrol note falls into.

Can I use it without API keys?

Yes, manual mode calls no model and costs nothing.

Where is data stored?

Only on the local computer.

Related projects

  • SiYuan

    Privacy-first personal knowledge management system, listed as integrating Jev for document classification and tag routing.

    PluginproductionTypeScript
  • docket

    Go terminal app that files documents using five Jev questions: category, sensitivity, urgency, whether action is needed and action state.

    ProjectbetaGo
  • jevault

    Obsidian plugin that ranks existing vault folders and tags for the active note with Jev and only moves or tags after you confirm.

    ProjectbetaTypeScript
  • LLM Wiki v2

    LLM-maintained Markdown wiki where a System One classifier decides where documents and questions go, keeping it fast as it grows.

    ServicebetaPython
  • paperless-jev

    Docker service that classifies the Paperless-ngx inbox with Jev (or local Clef via Ollama), setting confident fields and queueing the rest.

    Projectbeta
  • switchyard

    macOS menu-bar link router that opens each link in the right browser profile, asking Jev when no local rule matches and learning rules.

    ProjectbetaSwift