Guide · Agent skill
How to use the Release Announcement Writer agent skill
Updated · written with AI from the project's own documentation and checked against our catalogue data
The Release Announcement Writer is an agent skill: a packaged set of written instructions that an AI agent such as Claude loads when your request matches its purpose. This guide explains what the skill makes the agent do, how to install it from the ZIP on our page, how to ask for it in plain language, and what to check before publishing the output. By the end you will be able to get a channel-appropriate release announcement drafted for a changelog, blog, in-app note or social post.
Release Announcement Writerpaperclipai/paperclip · MIT Download the skill Quick start
- 1Download and unzip. Download the skill ZIP from this page and unzip it so you have a folder containing SKILL.md.
- 2Place the folder. Move it to ~/.claude/skills/ for all projects, or to the project's .claude/skills/ folder for one project.
- 3Or upload in the Claude apps. In the Claude apps, go to Settings → Capabilities → Skills and upload the skill.
- 4Describe the release. Give the version, the main changes, any breaking changes, and who reads the copy.
- 5Name the channel. Ask for a changelog entry, release notes, dev blog post, in-app note or social post, or ask for several.
- 6Review and verify. Check every number, link, date and migration step, and confirm the release has shipped before publishing.
What the skill makes the agent do
The skill gives the agent a fixed playbook for announcing software releases. Before writing, the agent asks who the reader is and picks a format: a terse changelog entry, detailed release notes, a dev blog post, an in-app message, or a social post. Each format has its own shape, so a changelog entry is not written like a blog post.
Every format follows the same opening order: what changed in the user's vocabulary, who it affects, and what the reader should do (migrate now, opt in, or no action needed). The skill also includes dated, explicit upgrade or migration steps and flags breaking changes with a **Breaking:** prefix.
- Asks who the reader is and selects the channel
- Leads with what changed, who it affects and what to do
- Adds upgrade or migration steps and flags breaking changes
- Avoids filler such as 'We are excited to announce'
Install the skill from the ZIP
Download the ZIP for this skill from our page and unzip it. The skill is delivered as a folder containing SKILL.md. To make it available to Claude Code on your machine, place the unzipped folder in ~/.claude/skills/. To make it available only inside one project, place it in that project's .claude/skills/ folder.
If you use the Claude apps instead, upload the skill under Settings → Capabilities → Skills.
- Personal use across projects: ~/.claude/skills/
- One project only: the project's .claude/skills/ folder
- Claude apps: Settings → Capabilities → Skills, then upload
Trigger the skill with natural requests
You do not need a special command. Describe the release and the surface you need, and the agent will match your request to the skill. Naming the channel helps, and so does giving the audience. The skill's own rule is to decide the audience before writing, so you may be asked a question first.
Good requests include the version, the main changes, and who reads the copy. Paste your changelog or PR list if you have one.
- 'Write a changelog entry for v2.3.0. The main change is CSV export, and the fix is a crash on large imports.'
- 'Draft a dev blog post for our API release. Include migration steps because the auth header format changed.'
- 'Write a one-sentence in-app note and a social post for the new dashboard, which is now generally available.'
Examples of what you get back
For a changelog request, expect a Markdown entry grouped under Added, Changed, Fixed, Deprecated and Breaking headings, with each line stating a user benefit. For a dev blog request, expect a 300 to 800 word draft with a hook, three to five what's-new bullets, an upgrade paragraph and a one-sentence look ahead.
Where the release affects users, the output includes explicit migration steps. Where a change is internal only, the skill is designed to say it should not be announced.
- Changelog: terse lines, each with a user-visible impact
- Release notes: adds a migration section, compatibility table and known issues
- Dev blog: hook, what's new, upgrade, what's next
- Social post: one sentence, one link, one image or clip
Check the output before you publish
The skill only drafts copy. A person should check every number, link and date, and confirm that the release has actually shipped. The skill's rules say to avoid promised dates for unshipped work and to avoid claims that cannot be verified, so remove any numbers you cannot back up.
The source also lists a post-publish checklist: the changelog should be in source control with the release, the blog date should match the ship date, all links should work, and the internal team should be notified before the public post goes live.
- Verify every number, link and date
- Confirm the feature has shipped, not only been merged
- Make sure breaking changes also appear in the upgrade guide
- Notify the internal team before publishing
Know what the skill runs and needs
The skill lists no requirements. It does not run scripts and does not use the network, so it only produces text in the conversation. It does not post anything to a channel.
The file also contains optional steps for recording release copy as structured records inside the Paperclip product. Those steps apply only when the Paperclip feature experimental.enableCases is enabled. If it is not enabled, ignore that part; the skill still writes the copy. If the Paperclip API returns '403 Cases are disabled', the skill is designed to report that and continue with the copy.
- Requirements: none listed
- Runs scripts: no
- Uses network: no
- Paperclip case recording is optional and needs experimental.enableCases enabled
What people use it for
- Writing a terse changelog entry grouped under Added, Changed, Fixed and Breaking
- Drafting a dev blog post with hook, what's new, upgrade and what's next sections
- Producing an in-app note with one sentence and one link
- Turning one release into a social post and a release-notes migration guide
Frequently asked questions
How do I install the release announcement agent skill?
Download the ZIP from this page and unzip it. Place the folder in ~/.claude/skills/ for personal use, or in a project's .claude/skills/ folder for one project. In the Claude apps, you can instead upload it under Settings → Capabilities → Skills.
Does this skill require anything else to run?
The skill lists no requirements. It does not run scripts and does not use the network, so it only drafts text in the conversation.
Is the release announcement skill free?
The source lists the item under the MIT licence. It does not list prices, so check the page for any terms before assuming anything beyond that.
Will the agent publish my release notes automatically?
No. The skill only drafts copy and does not post it to any channel. A person needs to review and publish the output.
Can it announce unfinished features?
The skill tells the agent not to write about work that is still in active development and to wait until it ships. It will not help pre-announce work in progress without human judgement.
What if my release has no user impact?
The skill says internal-only changes with no user impact should not be announced. Update internal docs instead.
New agent skills, repos and AI tools — every week
The best new skills and open-source projects, the fastest-growing repos, and one free prompt. Five minutes to read. Unsubscribe anytime.
Our verdict
This skill suits open-source or product teams, developer relations staff and writers who need consistent changelog, blog and migration copy across several channels. It is not suited to internal-only changes, unfinished features, or anyone expecting automatic publishing, since it only drafts text that a person must review and post.
Download the skill