generate-release-notes
Skill47 starsChanged 6 months agoMITView on GitHub
Generate comprehensive, categorized release notes between two Git refs (tags, branches, commits). Compares refs to gather commits and merged MRs, then produces polished release notes organized by category (Features, Bug Fixes, Improvements, Breaking Changes, Documentation). Works with any GitLab project accessible via the MCP server.
12 other projects use a version of this.
Adopted by
Most recently taken up by
- Azure/azure-functions-python-workertheir own version, first seen Oct 13, 2026
- open-edge-platform/edge-ai-librariestheir own version, first seen Oct 8, 2026
- openvinotoolkit/nncftheir own version, first seen Oct 8, 2026
- dotnet/vscode-csharptheir own version, first seen Oct 8, 2026
- tsukumijima/KonomiTVtheir own version, first seen Oct 8, 2026
- teambit/bittheir own version, first seen Oct 8, 2026
Other versions12 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 12
- teambit/bit★ 19ksame name
- redis/redis-py★ 14ksame name
- dotnet/sdk★ 3.2ksame name
- dotnet/vscode-csharp★ 3.1ksame name
- XRPLF/xrpl-dev-portal★ 2.3ksame name
- huytieu/COG-second-brain★ 1.3ksame name
- openvinotoolkit/nncf★ 1.2ksame name
- tsukumijima/KonomiTV★ 1.1ksame name
- openvinotoolkit/openvino.genai★ 615same name
- Azure/azure-functions-python-worker★ 357same name
2 more not listed
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository22 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- .github/copilot-instructions.mdCopilot instructions
- llms-full.txtllms.txt
- llms.txtllms.txt
- create-architectural-decision-recordSkill
- create-implementation-planSkill
- create-mcp-evaluationSkill
- create-mcp-toolSkill
- create-specificationSkill
- generate-project-documentationSkill
- git-commitSkill
- golang-patternsSkill
- golang-testingSkill
- go-safe-move-refactorSkill
- go-source-documentationSkill
- increase-test-coverageSkill
- modularize-go-packageSkill
- review-and-refactorSkill
- update-project-documentationSkill
- update-starlight-docsSkill
- upstream-contributionSkill
The file
We split this skill into the rules it gives an agent
4 rules. 0 are common, 4 are distinctive, and 0 are very specific to this project.
0
Common
Most of the 10 edited versions we read keep these.
4
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 129
Use past tense for descriptions ("Added", "Fixed", "Improved")
- Distinctive: line 130
Keep each entry to one line with the MR reference
- Distinctive: line 167
If the range spans hundreds of commits, the data may be truncated — the tool will indicate this
- Distinctive: line 168
Use specific tags rather than branch names for reproducible release notes
“Common” is measured: kept by at least 6 in 10 of the 10 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Generate Release Notes
Overview
Generate polished, categorized release notes for a GitLab project by comparing two Git refs. This skill orchestrates multiple MCP tools to gather commits, merged merge requests, and diffs, then produces human-readable release notes.
When to Use
- Creating release notes for a new tag or version
- Summarizing changes between any two Git references (tags, branches, SHAs)
- Preparing changelog entries for a release
- Reviewing what changed between deployments
Prerequisites
- A GitLab project accessible via the MCP server
- Two valid Git refs (tags, branches, or commit SHAs) to compare
Workflow
Step 1: Identify the Project and Refs
Determine the project ID (or namespace/project path) and the two refs to compare.
Common patterns:
- Tag to tag:
v1.0.0→v1.1.0 - Tag to branch:
v1.0.0→main - Branch to branch:
release/1.0→release/1.1 - Commit to tag:
abc1234→v1.1.0
Step 2: Gather Data (Choose One Approach)
Approach A: Manual Orchestration (Recommended)
Orchestrate the tools manually and categorize the results yourself. The tool names below are the meta surface (GITLAB_MCP_TOOL_SURFACE=meta); on the default dynamic surface the same calls go through gitlab_execute_action with the canonical action IDs repository.compare and merge_request.list, and on the individual surface the tools are gitlab_repository_compare and gitlab_mr_list:
Compare refs to get commits and diffs:
Use gitlab_repository with action "compare": - project_id: <project_id> - from: <base_ref> - to: <target_ref>Search for merged MRs in the change range:
Use gitlab_merge_request with action "list": - project_id: <project_id> - state: "merged" - per_page: 100- Correlate commits with MRs to enrich the data.
- Categorize changes based on MR labels, commit prefixes, and descriptions.
Approach B: Use the MCP Prompt
For a prompt-based approach that provides enriched context to the LLM:
Use the generate_release_notes prompt with:
- project_id: <project_id>
- from: <base_ref>
- to: <target_ref>Step 3: Format the Release Notes
Organize changes into these categories (skip empty categories):
## Release Notes: <from> → <to>
### Breaking Changes
- **!IID** — Description of breaking change (@author)
### Features
- **!IID** — New feature description (@author)
### Bug Fixes
- **!IID** — Bug fix description (@author)
### Improvements
- **!IID** — Improvement description (@author)
### Documentation
- **!IID** — Documentation change (@author)
### Other
- **!IID** — Other change description (@author)
---
**Full diff**: <compare_url>
**Contributors**: @user1, @user2, @user3
**Commits**: N | **Merge Requests**: M | **Files Changed**: FCategorization Rules
Assign each change to a category using these heuristics:
| Signal | Category |
Label breaking | Breaking Changes |
Commit prefix feat: | Features |
Label feature | Features |
Commit prefix fix: | Bug Fixes |
Label bug | Bug Fixes |
Commit prefix docs: | Documentation |
Label documentation | Documentation |
Commit prefix perf: | Improvements |
Commit prefix refactor: | Improvements |
Label enhancement | Improvements |
| Everything else | Other |
Priority: Labels take precedence over commit prefixes. Breaking changes always come first.
Writing Style
- Use past tense for descriptions ("Added", "Fixed", "Improved")
- Keep each entry to one line with the MR reference
- Include the author for attribution
- List Breaking Changes first — they need immediate attention
- Skip empty categories entirely
- Reference MR IIDs (e.g.,
!42) rather than commit SHAs for readability
Example Output
## Release Notes: v1.0.0 → v1.1.0
### Breaking Changes
- **!98** — Removed deprecated `/api/v1/users` endpoint (@alice)
### Features
- **!95** — Added webhook support for pipeline events (@bob)
- **!92** — Implemented project template cloning (@alice)
### Bug Fixes
- **!97** — Fixed race condition in concurrent MR updates (@carol)
- **!93** — Resolved TLS verification failure on self-signed certs (@bob)
### Improvements
- **!96** — Reduced API calls by caching project metadata (@carol)
- **!94** — Refactored pagination to use keyset pagination (@alice)
### Documentation
- **!91** — Updated installation guide for v1.1.0 (@bob)
---
**Full diff**: https://gitlab.example.com/group/project/-/compare/v1.0.0...v1.1.0
**Contributors**: @alice, @bob, @carol
**Commits**: 24 | **Merge Requests**: 8 | **Files Changed**: 37Tips
- If the range spans hundreds of commits, the data may be truncated — the tool will indicate this
- Use specific tags rather than branch names for reproducible release notes
- The prompt-based approach (Approach B) gives you the most control over the final output since you can edit the LLM response
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

