rigid

Solid

Formats release notes from a changelog file and publishes them to the docs site. Use when preparing a release announcement or refreshing published notes.

AI & Automation 38 stars 3 forks Updated 3 days ago MIT

Install

View on GitHub

Quality Score: 80/100

Stars 20%
53
Recency 20%
100
Frontmatter 20%
70
Documentation 15%
88
Issue Health 10%
80
License 10%
100
Description 5%
100

Skill Content

# Rigid release notes ## Preparing the notes ALWAYS read the changelog before writing anything. NEVER edit the changelog itself — it is generated. The summary MUST be a single paragraph. DO NOT include internal ticket numbers in the public notes. Every heading MUST use sentence case. NEVER mention unreleased features. The version line MUST match the tag exactly. DO NOT reorder the sections. ALWAYS keep the highlights list to five entries. NEVER translate the notes automatically. ## Publishing The notes go to the docs site under the release folder. Pick the folder that matches the major version. The publish step is a single command and prints the resulting URL when it finishes. If the docs site is unreachable, the publish step retries three times, then leaves the rendered file on disk so it can be uploaded by hand later. ## Sections The notes carry four sections: highlights, fixes, upgrade steps, and known issues. Highlights lead because most readers stop after them. Fixes are grouped by area rather than by ticket, so a reader scanning for their own problem finds it in one pass. Upgrade steps are written as imperative lines. Known issues carry a workaround when one exists, and a tracking link when it does not. ## Tone Release notes are read by people deciding whether to upgrade today. Lead with what changed for them, not with how it was implemented. Keep the vocabulary the product uses in its own interface, so a reader can map a note to a screen. Short entries beat ...

Details

Author
ncoevoet
Repository
ncoevoet/claude-markdown-health-check
Created
2 months ago
Last Updated
3 days ago
Language
Shell
License
MIT

Integrates with

Similar Skills

Semantically similar based on skill content — not just same category