JevMade field notes / Architecture and workflow guide

A guide to checking code for hidden mistakes using AI

You will learn how this project uses an AI tool to check computer code and documents for confusing mistakes. It explains how to set rules, test them, and review the results before making changes.

Original by mizchiEvaluation

Listen to this guide

JevMade’s plain-English explanation

0:00 /

AI narration

Credits and license

“jev-lint” by mizchi. Read the original source.

This expanded guide is an AI-narrated adaptation of the source’s essential explanation, examples and caveats, not a word-for-word reading. The synthetic voice does not imitate the author or imply their endorsement.

License: MIT

MIT License

Copyright (c) 2026 mizchi<Kotaro Chikuba>

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.

Our summary

This project is a tool that helps writers and programmers find confusing mistakes in their work. People use it to spot problems that standard checkers miss, like when a description says one thing but the actual instructions do something completely different.

The software searches for specific parts of a file and sends them to Jev, an AI tool that chooses from options rather than writing an answer. The AI scores how likely it is that a rule was broken, and flags the issue if the score is high enough.

This tool is useful for teams wanting to double-check their work, but it is not perfect. In the author's tests, about one in five alerts was wrong. Because the AI's scores can change slightly between runs, a person must always review the flagged issues before acting on them.

Key takeaways

Write specific rules that look for one clear mistake instead of general bad quality.
Record the AI's decisions so you can review them later when adjusting your alert settings.
Only ask the AI to double-check its answers if the results change too much between tests.

More guides →See the experiments →