A (dirty) bash script that will backup issues from the GitHub repo using API. It takes pagination into consideration. Originally created by @boombatower. Done under a hour, so it can be improved (grep anyone?), but hey, it's one time script and it works. What more do you need? :)

backup_github_issues.sh

#!/bin/bash

repo=""             # put the name of the repo here. Usually it is like this: organization/repo-name
username=""         # your github username
password=""         # your github password
numberofpages=      # leave blank for now and script will help you find the number of pages

if [ -z $numberofpages ]
  then
    echo "No number of pages set, lets find out how many pages are there"
    # -I will give back only the headers without the data
    curl -I -u "$username:$password" \
      "https://api.github.com/repos/$repo/issues?per_page=100&state=all"
    echo "Look above and find something like rel='last'. Just before that there will be the number of the last page. Put it in the variable numberofpages;"
  else
    echo "OK, all set, dumping issues..."
    echo
    echo

    for i in `eval echo {0..$numberofpages}`
    do
      filename=$(echo "$repo-issues-page$i.json" | tr / -)
      curl -u $username:$password \
        "https://api.github.com/repos/$repo/issues?per_page=100&state=all&filter=all&page=$i" \
        > $filename
      echo "Page $i finished..."
    done
fi

Works for me! Thanks 👍

This worked great but the issues were in reverse order. Anyone know an easy fix?

Cool, works well, thanks. A few notes:

• you can use this to dump a repo you don't own, as long as it's public access. I wasn't understanding this at first, since a password is needed...

• if you change everywhere it says issues to pulls, you get your Pull Requests :-)

In case anybody's interested... I just updated this on my fork with a version that handles number of pages automatically.