CSV to GitHub Issues importer

It creates issues in a repository, and GitHub only lets a signed-in user do that. You can drop and map a file without signing in; GitHub sign-in is asked for at the repository step. The rows go through Gitsu's server to GitHub. We keep no copy.

Title, body, labels (with the separator you pick), assignee, milestone, state, created date, closed date and comments. Missing labels and milestones are created in the repository first, and labels take the spelling the repository already uses. Every column you do not map is written at the end of the issue body, as a line such as "Sprint: 14", so nothing in the file is dropped. Each issue also ends with a line that marks it as imported from this file.

The issue is imported unassigned and the name from the file is written in the issue footer as "Assignee in CSV". The same happens when the login you matched cannot be assigned in the repository, for example when that person is not a collaborator. GitHub takes one assignee per imported issue.

GitHub works through one repository's import jobs about one at a time, roughly 70 a minute without comments. A file of 1,000 rows takes about 14 minutes, and the tab has to stay open for that long. A larger file is refused with its row count; split it and import the parts one after the other. The limit is a single setting, so it can change.