Stop writing release notes manually after every deployment. This workflow collects merged pull requests since the last release, uses AI to summarize changes into user-friendly release notes, and saves the documentation to Notion ready for publishing.
Detect new release in GitHub
The workflow triggers when a new release is published on GitHub. It captures the release tag, version number, and determines the previous release to identify the range of changes to document.
Fetch merged PRs since last release
The workflow queries GitHub for all pull requests merged between the previous release and the new one. It collects PR titles, descriptions, labels, authors, and linked issues to understand what changed.
Categorize and summarize changes with Claude
Using Claude, the workflow analyzes the PRs and categorizes them into sections like Features, Improvements, Bug Fixes, and Breaking Changes. The AI transforms technical PR descriptions into user-friendly summaries that explain the benefit of each change.
Generate formatted release notes with AI
The AI generates complete release notes with an introduction, categorized changes, contributor acknowledgments, and any upgrade instructions for breaking changes. The format follows your documentation style guide.
Save release notes to Notion
The generated release notes are saved to your Notion workspace as a new page in your changelog database. The page includes version number, date, and all sections ready for review and publishing.
Why automate release notes with AI?
Release notes are essential for communicating changes to users, but they're tedious to write and often get neglected or rushed. AI automation ensures every release gets comprehensive, user-friendly documentation without manual effort.
Never ship a release without documentation
When release notes are automatic, every release gets documented. No more blank changelog entries or delayed documentation that arrives days after the release.
Transform technical PRs into user-friendly content
PR descriptions are written for reviewers, not end users. AI translates technical changes into benefit-focused summaries that users actually care about.
Maintain consistent documentation quality
Manual release notes vary in quality depending on who writes them and how much time they have. AI produces consistent, well-structured documentation every time.
How to set up automated release notes
Setting up this GitHub to Notion workflow takes about 15 minutes. You'll connect your repository, configure categorization rules, and set up your Notion destination.
What you need to get started
- GitHub repository with release workflow
- Claude API key for content generation
- Notion workspace for documentation
- Defined release note categories and style
Configuring PR categorization
- Define how PR labels map to release note sections
- Specify which PRs to exclude (internal, dependencies)
- Set rules for identifying breaking changes
- Configure how to handle PRs without labels
Customizing release note format
- Provide your documentation style guide
- Define section order and headings
- Specify whether to include contributor names
- Configure any custom sections or metadata
Frequently asked questions about AI release notes
How does AI know what's important to users?
The AI analyzes PR titles, descriptions, and linked issues to understand the user impact. You can provide guidance about what types of changes matter most to your users.
Can I edit the generated notes before publishing?
Yes, notes are saved as drafts in Notion for review. Most teams do a quick review and light editing before publishing. The AI handles 80-90% of the work.
What if PRs have poor descriptions?
The AI does its best with available information. For better results, encourage good PR descriptions or use the AI to analyze the actual code changes for context.
Can this update our public changelog instead of Notion?
Yes, you can configure the workflow to update any documentation platform including your website, GitHub releases page, or dedicated changelog tools.