mcpbeat Sign in

Release Notes Style Agent Skill

Use when drafting or editing release notes or a changelog entry in the house voice.

6k tokens
context cost
the whole folder, loaded on every use
6
files
instructions only
0
copies elsewhere
how many repositories repackaged it
17
stars on the repo
on the repository, not the skill itself

Install

one command, takes just this skill from the repository
npx skills add https://github.com/vercel-labs/eve-content-agent-template --skill release-notes-style

What comes with it

22 244 bytes besides the instruction
references/ai-phrases-to-avoid.md
references/banned-words.json
references/best-practices.md
references/format-specs.md
references/plain-english-alternatives.md

The instruction itself

4 sections, as written by the author

Release notes voice & style

When writing or editing release notes:

  • Lead with what changed for the user, in their words — not the internal feature name. Say "you

can now…", not "we added…".

  • Group by what the reader does, not by team or component. Keep the same section names every

release so readers learn to scan it.

  • Each entry: what changed, why it matters, and (if relevant) what to do about it. Front-load the

benefit in the first sentence and keep the entry to a sentence or two.

  • Plain present tense: "You can now…", "Fixed an issue where…". Describe a fix from the symptom

the reader saw, not the code path. Skip jargon the reader wouldn't know.

  • Link to docs or a deeper post instead of explaining everything inline. Keep the entry scannable.
  • Be honest about breaking changes — call them out first, with the migration steps and a timeline

if there is one. Deprecate a feature for a release cycle before you remove it.

  • No marketing adjectives. A changelog is a record, not a pitch.
  • Every shipped version gets an entry, newest first, dated YYYY-MM-DD. Don't paste the commit

log — it's noise, not a record.

Structure

Group entries under clear headings in this order:

  • Breaking changes — what to do, up front (covers Changed/Deprecated/Removed/Security

that affect upgrades; name the migration path).

  • New — capabilities the reader didn't have before (Added).
  • Improved — things that already worked, now better (Changed).
  • Fixed — bugs resolved, described from the symptom the user saw (Fixed).

These map onto the Keep a Changelog category set (Added, Changed, Deprecated, Removed, Fixed,

Security). Use only the headings that have entries, and reach for the standard names rather than

inventing "Tweaks" or "Misc". See references/format-specs.md.

References

  • references/best-practices.md — sourced checklist of current release-notes conventions

(user-benefit framing, entry anatomy, grouping, breaking changes, linking).

  • references/format-specs.md — quick-lookup conventions: the Keep a Changelog category set,

SemVer mapping, entry-anatomy template, and ordering/dating norms.

  • references/banned-words.json — the list of words to avoid. Read it once up front for

awareness, then run the lint_against_style tool on your draft to check it against this

file before proposing the draft to the writer.

Shared references

Shared global references for all skills in the content agent. These apply regardless of the

content type.

  • references/ai-phrases-to-avoid.md — AI-tell words, phrases, and punctuation to avoid.
  • references/plain-english-alternatives.md — plain-English swaps for bloated or vague wording.

How to use it

Copy the folder

Take vercel-labs/release-notes-style from the repository into ~/.claude/skills for personal use, or into .claude/skills inside a project.

Check the name does not clash

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.