github-milestones
Skill0 starsChanged 3 months agoMITView on GitHub
How the user's GitHub milestones work — bucket milestones (Critical, Needed soon, Needed later, Not urgent, Upstream) versus release and deadline milestones, when to put an issue on one, how a milestone becomes a release, and when to suggest moving a due date. Use before setting or changing an issue's milestone, creating, editing or closing a milestone, planning or cutting a release, or choosing what to work on next from a repository's milestones.
In this repository9 other files it gives its agents.
- CLAUDE.mdCLAUDE.md
- lookup-mesh-tree-numbers/CLAUDE.mdCLAUDE.md
- settings/CLAUDE.mdCLAUDE.md
- skills/CLAUDE.mdCLAUDE.md
- copilot-reviewSkill
- github-issuesSkill
- pr-updateSkill
- sync-docsSkill
- wrapSkill
The file
We split this skill into the rules it gives an agent
2 rules. About 0 are common, 2 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈2
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 32
If it can't be fetched, say so and leave buckets alone rather than working from the names.
- Distinctive: line 55
Every issue should end up on a milestone — a bucket, or a release it is needed for — but that is the user's triage to do, not yours.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
github-milestones
Milestones come in two kinds, and nearly every mistake with them is treating one kind as the other: a bucket read as a deadline, or a soft release date read as an alarm.
These are the user's conventions, for repositories where they run the milestones. A repository that
documents its own scheme — or says to track releases in a GitHub Project — in its CLAUDE.md,
CONTRIBUTING.md or README follows that instead.
See what a repository has
gh api "repos/<owner>/<repo>/milestones?state=open" --paginate \
--jq '.[] | "\(.number)\t\(.title)\t\(.due_on // "undated")\t\(.open_issues) open"'Buckets
The buckets, what each one means, and the rules for them live in MILESTONES.md in the
gaurav/milestones repository. Read it before putting an issue in a bucket or touching a bucket
milestone:
gh api repos/gaurav/milestones/contents/MILESTONES.md --jq .content | base64 -dIf it can't be fetched, say so and leave buckets alone rather than working from the names.
Release and deadline milestones
Named for a release (tool-name v1.2) or a date (2026aug22). These are usually dated, and the
date means one of two things:
- Hard or semi-hard — a delivery expected for testing on that date, or the last day a build can start and still finish before another deadline.
- Soft — in a repository whose last release was v1.1,
v1.2can simply mean the next release, whenever that is: ranked above Needed soon, with a date that is a hope.
Most are soft, so most drift into the past. A soft milestone being past due is normal and not something to report.
Milestones are releases
Unless the repository says otherwise, a release milestone's issues are what the release contains. The release is built from them and published as a GitHub release, after which the milestone is closed. Closing it is part of that outward-facing step: ask first, as for the release itself.
Putting an issue on a milestone
Every issue should end up on a milestone — a bucket, or a release it is needed for — but that is the user's triage to do, not yours. Leave a new or untouched issue unmilestoned unless it is clearly Critical or clearly needed for a particular release — the user's triage workflow picks up anything without a milestone. A guessed bucket hides an issue from that workflow, which is worse than no bucket.
gh issue edit <N> --repo <owner>/<repo> --milestone "<title>"
gh issue create --repo <owner>/<repo> --milestone "<title>" ...Whose issue it is doesn't matter here: setting a milestone is triage, which the github-issues
skill allows on anyone's issue.
Nudging about a due date
Suggest a due-date change and let the user make it; don't edit the milestone yourself. A date being in the past is not by itself worth mentioning — most are soft. What is worth mentioning:
- A milestone really is urgent and its date doesn't show it: a hard deadline mentioned in an issue or in conversation, Critical issues accumulating ahead of a release, an upstream contact that has to happen soon. Say which milestone, what date, and what the evidence was.
- A hard deadline is at risk — close or already passed, with issues still open.
- An urgency date has outlived the urgency — a bucket still carrying a deadline that has passed and no longer matters. Suggest removing it.
This is about due dates only. Anything else that looks wrong with a milestone, raise as you would anywhere else.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.