Programming · Writing
writing-changesets
/writing-changesets
Use when writing a changeset, or before filing a PR in a repository that uses changesets.
For whom
- Author
- adelrodriguez ↗
- Repository
- skills ↗
- Licence
- MIT
- Last repository change
- 11.11.2026
Assessment and security
100/ 100Class: good
Format25 / 25
Frontmatter with name and description, valid name, description up to 1,024 characters, name matching the folder.
Discoverability25 / 25
Description explains when to use the skill, has a reasonable length and uses third person.
Content35 / 35
At least 150 words and at most 500 lines, headings, step lists, examples or code.
Hygiene15 / 15
No placeholders such as TODO, no Windows paths, with a licence.
Riskno signals detected in the scan
- Files in this package
- 1
- Scripts
- 0
- HTML comments
- 0
- URLs in SKILL.md
- 0
Deprecation
No signals detected. Repository changed on 11.11.2026, without references to retired models or APIs.
Licence
MIT, open: use and modify while retaining author attribution.
What it does and when to use it
Use when writing a changeset, or before filing a PR in a repository that uses changesets.
Original author description
Use when writing a changeset, or before filing a PR in a repository that uses changesets.
How it works
- Find the change
- Sort the changes
- Match the pending changesets
- Pick the version bump
- Write each changeset
- Check the changesets
- Return the result
Original SKILL.md headings · 86 lines, 734 words
SKILL.md preview
--- name: writing-changesets description: Use when writing a changeset, or before filing a PR in a repository that uses changesets. --- # Write changesets Write the changesets for a diff, so the release notes tell the developers who use the package what changed for them. A changeset is one entry file in `.changeset/`. A change is one feature, fix, or breaking change, not one file. ## Find the change Read the staged changes with `git diff --staged`. When nothing is staged, read the branch's changes against its base with `git diff origin/<base>...HEAD`. Read `.changeset/config.json` for packages it ignores and packages it versions together. Only packages that are published, which means not `private` in their `package.json`, get a changeset. The step is done when you have the diff and the list of published packages it touches. ## Sort the changes Put each change in the diff in one of these groups: - **Write.** A developer using the package can notice it: a new feature, a fixed bug, a breaking change, a performance gain, or a dependency update that changes behavior. - **Skip.** Internal refactoring, reorganized code, test-only changes, and cosmetic cleanups. When internal cleanup carries a small user-facing fix, put only the fix in the write group. When the write group is empty, write no changeset, and say so. The step is done when every change is in exactly one group. ## Match the pending changesets Read every changeset in `.changeset/`, ignoring `README.md`. These are pending: their changes are not released yet. For each change in the write group, decide: - **Expand.** A pending changeset covers the same feature, fix, or API in the same package. When the change fixes or adjusts something that changeset added, fold it in without a separate mention, because users never saw the earlier version. - **New.** No pending changeset covers it. Give it its own changeset, even when another change in the same package also gets one. The step is done when every change in the write group is marked expand, with the changeset's path, or new. ## Pick the version bump Use semantic versioning, with these choices: - Bump a performance gain or a dependency update as a patch. - Below 1.0.0, bump a breaking change as a minor. Use major only when the user asks for it. - When you expand a changeset, keep the higher of its bump and the change's bump. The step is done when every changeset you will write has a bump for each package it names. ## Write each changeset Rewrite each expanded changeset in place, so it describes the combined result. Create each new changeset with the repository's package runner, such as `bunx changeset --empty`, and fill in the file it creates. A changeset is reference: it describes what changed, in the product's terms. Follow `references/reference.md` and `references/style.md` in the `writing-documentation` skill. For a breaking change, add the steps to migrate as a short how-to, following `references/how-to-guide.md` in the same skill. Start with one line that names the change from the user's side: - Avoid: `This changeset updates the internal implementation to improve code organization` - Avoid: `The system has been enhanced to provide improved functionality` - Prefer: `Fix race condition in file watcher when multiple saves occur rapidly` - Prefer: `Add --dry-run flag to preview changes without writing files` Add a paragraph only when a user needs more than that line to use the change, such as how a new option behaves or how to migrate: ```markdown --- "package-name": minor --- Add `--watch` mode for continuous file monitoring Runs the command once, then re-runs whenever source files change. Uses filesystem events when available, and falls back to polling on unsupported platforms. ``` Use the `unslop` skill on the prose. The step is done when every change in the write group is in a changeset. ## Check the changesets - Every change in the write group is in exactly one changeset, and no changeset mentions a change in the skip group. - Every change that a pending changeset covers expanded it instead of adding a new one. - Every changeset names only published packages that its changes touch. - Every breaking change has a major bump, or a minor bump below 1.0.0, and steps to migrate. - Every changeset's first line names the change from the user's side. ## Return the result Return the path of each changeset you created or expanded, marked as new or expanded, with its packages, bumps, and first line. List the changes you skipped, one line each.
Versions
The database has no other skills named writing-changesets.
Database snapshot: 13.11.2026 · package version b6b181b05782 (SHA-256 digest of the entire package)
Downloads in the last 30 days: 0 · ranking
Report a problem with this skill: harmful content, incorrect descriptions or an author’s removal request.
Similar skills
- 96changeset
Use when writing changesets for releases - 100changesets
Create or refresh a `.changeset/<slug>.md` for the current branch, or report that none is required. Triggers on "/changesets create", "add a changeset", "create a changeset", "update the changeset", "refresh the changeset", "do I need a changeset", or any work that touches… - 100changeset-pr
Create or update a `.changeset/*.md` file for the current branch or PR in this repository, choose the correct package scope and release type, and verify the result against repo-specific Changesets config. Use when a publishable package changed, when a PR is missing a changeset,… - 90changesets
Concise PR finalization changesets. Use when implementation scope is complete and focused validation has finished to classify release impact, consolidate the PR's existing changeset, or write its consumer release note. - 100changesets
Write the changeset for a change in a repo that releases with changesets (a `.changeset/` folder exists). The bump level follows one question, whether the thing existed before: patch for changed behavior, minor for something new, major only with the maintainer's approval. Use… - 100changeset
Writes Changesets entries for a JavaScript or TypeScript monorepo. Detects the workspaces from pnpm-workspace.yaml or the package.json workspaces field, finds the packages changed on the branch, skips the ones ignored in .changeset/config.json, picks patch, minor or major for…