gh-issue-creator

name	gh-issue-creator
description	Create GitHub subject-tracking issues for this masters repository by using the lfariabr/gh-issue-creator repository. Use when asked to create module issues, assessment issues, subject epics, labels, or milestones based on a subject README or recent closed issue patterns.

GH Issue Creator

Use this skill to create subject issue sets in lfariabr/masters-swe-ai in the same style as prior terms, while routing batch issue creation through https://github.com/lfariabr/gh-issue-creator.

Required Tool

Treat the GitHub repository as the canonical source:

https://github.com/lfariabr/gh-issue-creator

Use one of these setup paths:

git clone https://github.com/lfariabr/gh-issue-creator /private/tmp/gh-issue-creator
python3 /private/tmp/gh-issue-creator/issue_creator.py --help

or download only the script:

curl -fsSL -o /private/tmp/issue_creator.py https://raw.githubusercontent.com/lfariabr/gh-issue-creator/main/issue_creator.py
python3 /private/tmp/issue_creator.py --help

If a clone already exists in a temporary path, update it with git -C /private/tmp/gh-issue-creator pull before use. Do not hardcode a user-machine-specific path.

Workflow

1. Resolve repository and subject context.

   • Confirm the GitHub repo from git remote -v or gh repo view.
   • Read the subject README and supporting notes if needed.
   • Check for existing issues before creating anything:

     gh issue list --state all --limit 1000 --search <SUBJECT_CODE> --json number,title,state,url
     

2. Copy the current repo pattern.

   • Inspect recent closed issues:

     gh issue list --state closed --limit 30 --json number,title,body,labels,milestone,closedAt,url
     

   • In this repo, subject issue sets normally use:
     • [Epic] <SUBJECT>-Modules
     • [Epic] <SUBJECT>-Assessments
     • Module N: <module title>
     • Assessment N: <type> - <title>
     • labels: subject label plus learning, assessment, and epic
     • assessment milestones attached only to assessment issues

3. Prepare labels and milestones before the batch create.

   • gh-issue-creator can attach labels and milestones, but it does not create them.
   • Check labels:

     gh label list --limit 100 --json name,color,description
     

   • Create the subject label if missing. Use a distinct readable color and description:

     gh label create <SUBJECT_CODE> --color <HEX> --description "<Subject name> subject"
     

   • Check milestones:

     gh api --method GET 'repos/<owner>/<repo>/milestones?state=all&per_page=100'
     

   • Create subject-prefixed assessment milestones if missing:

     gh api repos/<owner>/<repo>/milestones -f title='<SUBJECT> Assessment 1 Due' -f description='<deliverable>' -f due_on='YYYY-MM-DDT00:00:00Z'
     

   • Verify due dates after creation because GitHub timezone display can be surprising.

4. Generate a JSON template for gh-issue-creator.

   • Write the template outside the repo by default, for example /private/tmp/<subject>-issues.json, unless the user wants a committed artifact.
   • Include all epics, module issues, and assessment issues in one JSON template.
   • Use placeholder epic bodies if child issue numbers are not known yet, then update epics after creation.
   • Use milestone titles, not numbers, in the JSON template because gh issue create --milestone accepts milestone names.
   • Example issue object:

     {
       "title": "Module 1: Introduction to Big Data and Analytics",
       "body": "**Week 1** - Introduction to Big Data and Analytics\n\n**Learning Outcomes:** SLO a\n\n**Parent Epic:** #TBD\n\n---\nPart of BDA601 Big Data and Analytics learning modules.",
       "labels": ["learning", "BDA601"],
       "assignees": [],
       "milestone": null
     }
     

5. Dry-run, inspect, then create.

   • Always run a dry-run first:

     python3 /private/tmp/gh-issue-creator/issue_creator.py --template /private/tmp/<subject>-issues.json --repo <owner>/<repo>
     

   • If the user already asked to create the issues and the dry-run exactly matches the requested scope, proceed with:

     python3 /private/tmp/gh-issue-creator/issue_creator.py --template /private/tmp/<subject>-issues.json --repo <owner>/<repo> --create
     

   • If the dry-run reveals duplicates, missing labels, missing milestones, or an unexpected issue count, stop and resolve that before creating.

6. Update parent epics after creation.

   • List the new subject issues:

     gh issue list --state open --label <SUBJECT_CODE> --limit 50 --json number,title,labels,milestone,url
     

   • Update the module epic with actual module issue numbers.
   • Update the assessment epic with actual assessment issue numbers, due dates, weights, dependencies, and milestone names.
   • Use gh issue edit --body-file or the GitHub connector if available.

7. Verify and report.

   • Verify the final issue set:

     gh issue list --state open --label <SUBJECT_CODE> --limit 50 --json number,title,labels,milestone,url
     

   • Confirm:
     • expected issue count
     • subject label on every issue
     • learning labels on module issues and module epic
     • assessment labels on assessment issues and assessment epic
     • epic label on both epics
     • assessment milestones attached to assessment issues
   • Final response should list created epics, module issue range, assessment issues, labels, milestones, and any assumptions.