JevMade

Sign in
← Back to experiments

Apps & data pipelines

GitHub Issue Classifier

Jennifer Marsman's issue-labeling app asks Microsoft-Decision-1 for suggestions that a person can inspect and edit before a separate GitHub publication step.

Source screenshot of GitHub Issue Classifier
SOURCE SCREENSHOTFull screenshot ↗

What it does

It classifies open and closed issues with Choice questions, using server-side model credentials. Publishing requires a fine-grained GitHub token and skips assignments below the default 75% confidence cutoff, including manually changed labels. Despite the TypeSafe-shaped configuration names, the model is Microsoft-Decision-1, not Jev.

How you can use it

Define your issue labels, inspect the suggestions and correct individual assignments before publishing. Public repositories can be read without a GitHub token. Reading a private repository or writing labels needs the documented permissions. A manual edit does not bypass the default confidence rule for publication.

Keep this for later

Sign in to bookmark experiments, guides and videos, and keep notes only you can see.

Continue to sign in

We’ll bring you back to this listing.