Communication

commit-message-storyteller - Claude MCP Skill

Analyzes git diffs or staged changes and generates narrative commit messages that explain WHY a change was made, not just what changed — following Conventional Commits format. Use when asked to "write a commit message", "generate a commit", "describe my changes", "what should I commit this as", "commit this", "summarize my diff", or "help me commit". Works with git diff output, staged files, or plain descriptions of changes.

SEO Guide: Enhance your AI agent with the commit-message-storyteller tool. This Model Context Protocol (MCP) server allows Claude Desktop and other LLMs to analyzes git diffs or staged changes and generates narrative commit messages that explain why a chan... Download and configure this skill to unlock new capabilities for your AI workflow.

🌟1 stars • 4885 forks
📥0 downloads

Documentation

SKILL.md
# Commit Message Storyteller

Transforms raw git diffs and change descriptions into clear, story-driven commit messages that follow the [Conventional Commits](https://www.conventionalcommits.org/) specification. Instead of "update file.js", you get messages that communicate intent, context, and impact.

## When to Use This Skill

- User says "write a commit message", "help me commit", or "generate a commit"
- User pastes a git diff or describes code changes
- User says "what should I commit this as?" or "summarize my diff"
- User wants better commit history for their team or open-source project
- User is preparing a pull request and wants meaningful commit messages

## Prerequisites

Have at least one of the following ready:
- Output from `git diff` or `git diff --staged`
- A description of what you changed and why
- A list of modified files

## How It Works

### Step 1: Gather the Change Context

Ask the user (or infer from the diff) for:

1. **What changed** — files, functions, logic affected
2. **Why it changed** — bug fix, new feature, refactor, performance, etc.
3. **Who/what triggered it** — issue number, user request, tech debt, etc.

If the user provides a raw `git diff`, extract this context automatically from the diff.

### Step 2: Identify the Commit Type

Map the change to a Conventional Commits type using this guide:

| Type | Use When |
|------|----------|
| `feat` | A new feature or capability is added |
| `fix` | A bug or incorrect behavior is corrected |
| `refactor` | Code restructured without changing behavior |
| `perf` | A change that improves performance |
| `docs` | Documentation only changes |
| `style` | Formatting, whitespace, missing semicolons (no logic change) |
| `test` | Adding or updating tests |
| `chore` | Build process, dependency updates, config changes |
| `ci` | CI/CD pipeline changes |
| `revert` | Reverting a previous commit |

See `references/conventional-commits-guide.md` for detailed examples.

### Step 3: Write the Commit Message

Follow this structure:

```
<type>(<optional scope>): <short imperative summary>

<body — the story: why this change was made, what problem it solves>

<footer — issue refs, breaking change notices>
```

#### Rules for Each Part

**Subject line (first line):**
- Use imperative mood: "add", "fix", "remove" — not "added" or "fixes"
- Max 72 characters
- No period at the end
- Lowercase after the colon

**Body (the story):**
- Explain the *why*, not the *what* (the diff already shows the what)
- Describe the problem that existed before this change
- Mention any alternatives considered if relevant
- Keep lines under 100 characters
- Separate from subject with a blank line

**Footer:**
- Reference issues: `Closes #123`, `Fixes #456`, `Refs #789`
- Mark breaking changes: `BREAKING CHANGE: <description>`

### Step 4: Generate Output

Produce the commit message in a copyable code block, followed by a one-line plain-English explanation of the story you told.

**Example output:**

```
fix(auth): prevent token refresh loop on expired sessions

When a user's session expired mid-request, the auth middleware was
triggering a token refresh, which itself failed validation and triggered
another refresh — causing an infinite retry loop that crashed the app.

This adds a recursion guard flag that aborts the refresh cycle if a
refresh is already in progress, returning a clean 401 instead.

Closes #312
```

> **Story told:** A silent infinite loop on session expiry was crashing the app; this stops the cycle early and returns a clean error.

---

## Multiple Commits from One Diff

If the diff contains **logically separate changes**, split them into multiple commit messages and tell the user. Use this heuristic:

- Different files with unrelated purposes → likely separate commits
- Same file but distinct concerns (e.g., bug fix + refactor) → suggest splitting
- Everything tightly coupled → one commit is fine

---

## Edge Cases

| Situation | How to Handle |
|-----------|---------------|
| User provides no context beyond a diff | Infer type and scope from file names and changed symbols |
| Changes span many files with no clear theme | Ask: "Is this one logical change, or multiple?" |
| Breaking change detected | Add `BREAKING CHANGE:` footer automatically |
| User says "keep it short" | Omit body, just write a strong subject line |
| No issue number available | Omit the footer entirely |

---

## Quick Reference

```bash
# Get your staged diff to paste into Copilot
git diff --staged

# Or get the last uncommitted working tree changes
git diff
```

See `references/conventional-commits-guide.md` for type examples and scope guidelines.

Signals

Avg rating0.0
Reviews0
Favorites0

Information

Repository
github/awesome-copilot
Author
github
Last Sync
9/5/2026
Repo Updated
9/5/2026
Created
5/4/2026

Reviews (0)

No reviews yet. Be the first to review this skill!

Related Skills

firecrawl-build-search

Integrate Firecrawl `/search` into product code and agent workflows. Use when an app needs discovery before extraction, when the feature starts with a query instead of a URL, or when the system should search the web and optionally hydrate result content.

35092

firecrawl-build-onboarding

Get Firecrawl credentials and SDK setup into a project. Use when an application needs `FIRECRAWL_API_KEY`, when an agent should add Firecrawl to `.env`, when the user wants to authenticate Firecrawl for app code, or when choosing the first SDK and docs for a new Firecrawl integration. This skill includes its own browser auth flow, so it does not depend on the website onboarding skill.

35092

firecrawl-build

Integrate Firecrawl into application code whenever a product, agent, or workflow needs web data inside the app — web search, live search results, page scraping, structured extraction, or browser interaction. Use when building any feature that needs data from the web in code, even if the user does not mention Firecrawl explicitly and only describes wanting web data, website content, search, scraping, or interaction in an application. Trigger for Firecrawl requests, "fire girl" shorthand, and generic app-level web-data needs that should map to `/scrape`, `/search`, or `/interact`. Do not use this skill for one-off terminal-only web tasks during the current session; use `firecrawl/cli` for those.

35092

firecrawl-build-interact

Integrate Firecrawl `/interact` into product code for dynamic pages and browser actions after scraping. Use when a feature needs clicks, form fills, pagination, authentication-aware flows, or other multi-step interactions that plain `/scrape` cannot complete.

35092

Related Guides