How to compose or revise the weekly digest email of a repository's open issues: subject line, grouping, one-line issue summaries, needs-attention and stale criteria, citation format, and the closing invitation to reply. Load when composing this week's digest or reworking a draft of it. Not for other emails, replies, or GitHub and Linear comments.
npx skills add https://github.com/vercel-labs/kody-eve-template --skill digest-format
The weekly digest is one email a maintainer skims in their inbox. Everything here serves that: short, grouped, every issue cited as #N with a link, and a clear way to act by replying.
Open with a one-paragraph overview: total open issues, what changed since last week (new, closed, spikes), and any theme worth a sentence. Then the groups, then the closing invitation.
Default grouping, in this order:
When the user prefers a different grouping (by label, by assignee, by milestone), use theirs and keep the needs-attention issues at the top of whatever group they land in, marked as such.
Skip any group that would be empty rather than showing an empty heading. If the repo has no open issues at all, send a short note saying so instead of the grouped digest.
Needs attention, any of:
Stale: no comments, label changes, or other activity in 30 or more days. Staleness is about silence, not age; an old issue with fresh discussion belongs in recent activity.
An issue appears in exactly one group. Needs attention wins over the others.
Each issue gets one line: the citation, the title or a tightened version of it, then the state of play.
Example line: #42 Crash on empty config: fix proposed, awaiting maintainer review.
End every digest by inviting the reader to reply to act, naming the concrete options: ask for detail on an issue, have you comment on one, or have you create Linear issues from it. Reference a real issue number from this digest in the example so the reply pattern is obvious, for example "reply with 'create Linear issues for #12 and #17'".
Take vercel-labs/digest-format from the repository into ~/.claude/skills for personal
use, or into .claude/skills inside a project.
The agent identifies a skill by the name field in its header. Two skills with the
same name cannot sit side by side — one of them will be ignored.