---

name: manage-changelog locale: caveman-ultra source_locale: en source_commit: 82c77053 fence_basis_commit: 82c77053 translator: "Julius Brussee homage — caveman" translation_date: "2026-04-24" description: > Maintain changelog per Keep a Changelog. Entry categorization (Added, Changed, Deprecated, Removed, Fixed, Security), version section mgmt, Unreleased tracking. Use starting new project needing changelog, adding entries after features/fixes, prepping release by promoting Unreleased → versioned section, or converting free-form → Keep a Changelog format. license: MIT allowed-tools: Read Write Edit Grep metadata: author: Philipp Thoss version: "1.0" domain: versioning complexity: basic language: multi tags: versioning, changelog, documentation, keep-a-changelog

Manage Changelog

Maintain project changelog per Keep a Changelog. Create new, categorize entries, manage [Unreleased], promote to versioned on release. Adapts R convention (NEWS.md) when detected.

Use When

• Start new project needing changelog
• Add entries after features / fixes / other changes
• Prep release → move Unreleased → versioned section
• Review completeness before publish
• Convert free-form → Keep a Changelog format

In

• Req: Project root dir
• Req: Change description (or git log to extract from)
• Opt: Target version # (release promotion)
• Opt: Release date (default today)
• Opt: Format pref (Keep a Changelog / R NEWS.md)

Do

Step 1: Locate / Create Changelog

Search existing changelog in project root.

# Check for common changelog filenames
ls -1 CHANGELOG.md CHANGELOG NEWS.md CHANGES.md HISTORY.md 2>/dev/null

None exists → create w/ std header:

# Changelog

All notable changes to this project will be documented in this file.

The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/),
and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).

## [Unreleased]

R pkgs → use NEWS.md w/ R convention:

# packagename (development version)

## New features

## Bug fixes

## Minor improvements and fixes

→ Changelog located / created w/ proper header + Unreleased section.

If err: Exists in non-std format → don't overwrite. Note format diff + adapt entries to match existing style.

Step 2: Parse Existing Entries

Read + ID structure:

1. Header/preamble (project name, format desc)
2. [Unreleased] w/ pending changes
3. Versioned in reverse chron order ([1.2.0] before [1.1.0])
4. Comparison links at bottom (opt)

Per section → ID categories:

• Added — new features
• Changed — changes in existing fn
• Deprecated — soon-to-be removed
• Removed — now removed
• Fixed — bug fixes
• Security — vulnerability fixes

→ Structure understood, existing entries inventoried.

If err: Malformed (missing sections, wrong order) → note issues, don't restructure w/o confirmation. Add new correctly + flag structural issues for manual review.

Step 3: Categorize New Changes

Per change → classify into 1 of 6:

Category	When to Use	Example Entry
Added	New feature or capability	- Add CSV export for summary reports
Changed	Modification to existing feature	- Change default timeout from 30s to 60s
Deprecated	Feature marked for future removal	- Deprecate old_function()in favor ofnew_function()``
Removed	Feature or capability removed	- Remove legacy XML parser
Fixed	Bug fix	- Fix off-by-one error in pagination
Security	Vulnerability fix	- Fix SQL injection in user search (CVE-2026-1234)

Entry writing guidelines:

• Start w/ imperative verb (Add, Change, Fix, Remove)
• Specific enough user understands impact w/o reading code
• Ref issue #s / CVEs where applicable
• One line; sub-bullets only for complex changes

→ Each change in exactly one category w/ well-written entry.

If err: Change spans multi categories (feature + bug fix) → separate entries per category. Unclear → default "Changed".

Step 4: Add Entries to Unreleased

Insert under [Unreleased]. Maintain category order: Added, Changed, Deprecated, Removed, Fixed, Security.

## [Unreleased]

### Added

- Add batch processing mode for large datasets
- Add `--dry-run` flag to preview changes without applying

### Fixed

- Fix memory leak when processing files over 1GB
- Fix incorrect timezone handling in date parsing

Only add categories w/ entries; don't include empty category headings.

→ New entries added under [Unreleased] in correct categories, consistent formatting.

If err: Unreleased section missing → create immediately below header/preamble + above first versioned section.

Step 5: Promote to Versioned on Release

Release cut → move Unreleased → new versioned section:

1. New heading: ## [1.3.0] - 2026-02-17
2. Move all entries from [Unreleased] → new section
3. Leave [Unreleased] empty (keep heading)
4. Update comparison links at bottom

## [Unreleased]

## [1.3.0] - 2026-02-17

### Added

- Add batch processing mode for large datasets

### Fixed

- Fix memory leak when processing files over 1GB

## [1.2.0] - 2026-01-15

### Added

- Add CSV export for summary reports

Update comparison links (if present):

[Unreleased]: https://github.com/user/repo/compare/v1.3.0...HEAD
[1.3.0]: https://github.com/user/repo/compare/v1.2.0...v1.3.0
[1.2.0]: https://github.com/user/repo/compare/v1.1.0...v1.2.0

R NEWS.md → use R convention:

# packagename 1.3.0

## New features

- Add batch processing mode for large datasets

## Bug fixes

- Fix memory leak when processing files over 1GB

# packagename 1.2.0
...

→ Unreleased entries moved to dated versioned section; Unreleased cleared; comparison links updated.

If err: Version # conflicts w/ existing → already released. Check apply-semantic-versioning for correct version.

Step 6: Validate Format

Verify meets format requirements:

1. Versions in reverse chron (newest first)
2. Dates ISO 8601 (YYYY-MM-DD)
3. Each versioned has ≥1 categorized entry
4. No dup version sections
5. Comparison links match version sections

# Check for duplicate version sections
grep "^## \[" CHANGELOG.md | sort | uniq -d

# Verify date format
grep "^## \[" CHANGELOG.md | grep -v "Unreleased" | grep -vE "\d{4}-\d{2}-\d{2}"

→ Passes all format checks w/ no warnings.

If err: Fix format issues: reorder, correct dates, remove dups. Report issues requiring human judgment (missing entries for known changes).

Check

• File exists w/ proper header ref Keep a Changelog + SemVer
• [Unreleased] at top (below header)
• All new entries categorized into Added/Changed/Deprecated/Removed/Fixed/Security
• Entries start w/ imperative verb + describe user-facing impact
• Versioned sections reverse chron
• Dates ISO 8601 (YYYY-MM-DD)
• No dup version sections
• Comparison links (if used) correct + up to date
• Empty categories not included (no heading w/o entries)

Traps

• Internal-only entries: "Refactored DB module" not useful to users. Focus on user-facing. Internal refactors → commit msgs, not changelogs.
• Vague entries: "Various bug fixes" tells user nothing. Each fix → specific descriptive entry.
• Forget Unreleased: Adding directly to versioned = changes documented as released when not.
• Wrong category: "Fix" actually adds feature. Fix restores expected behavior; new capability = "Added" even if requested as bug report.
• Missing Security: Security fixes always documented w/ CVE ids when avail. Users need to know if should upgrade urgently.
• Changelog drift: Not updating at time of change. Batch-writing before release → missed / poorly described changes. Write entries alongside code changes.

→

• apply-semantic-versioning — determine version # pairing w/ entries
• plan-release-cycle — define when entries promoted to versioned
• commit-changes — commit changelog updates w/ proper msgs
• release-package-version — R-specific release workflow incl NEWS.md updates
• create-github-release — use changelog content as GitHub release notes