# Summarize Changes

> Summarizes uncommitted git changes and flags anything risky. Use when the user asks what changed, wants a commit message draft, or asks to review their diff before committing.

- **Type:** Skill
- **Install:** `agentstack add skill-davila7-claude-with-skills-summarize-changes`
- **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior
- **Seller:** [davila7](https://agentstack.voostack.com/s/davila7)
- **Installs:** 0
- **Category:** [Agent Skills](https://agentstack.voostack.com/c/agent-skills)
- **Latest version:** 0.1.0
- **License:** MIT
- **Upstream author:** [davila7](https://github.com/davila7)
- **Source:** https://github.com/davila7/claude-with-skills/tree/main/src/content/docs/01-basic/lesson-05-invoking/examples/summarize-changes
- **Website:** https://claude-with-skills.vercel.app

## Install

```sh
agentstack add skill-davila7-claude-with-skills-summarize-changes
```

Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent.

## About

## Current changes

!`git diff HEAD`

## Instructions

Summarize the changes above in two or three bullet points. Focus on the logical intent of the changes, not the line-by-line diff.

Then list any risks you observe, such as:

- Missing error handling on new code paths
- Hardcoded values (API URLs, credentials, magic numbers) that should be configuration
- Removed tests or disabled assertions
- Changes to public API signatures or exported interfaces
- Large refactors mixed with behavior changes (should be separate commits)

If the diff is empty, say there are no uncommitted changes.

---

Note for learners: the `!` backtick line above is dynamic context injection. When Claude Code activates this skill, it runs `git diff HEAD` and replaces that line with the real diff output before Claude reads the file. The result is that Claude sees the actual diff as part of the skill body, without needing a separate tool call. This is the canonical example from the Claude Code skills documentation.

## Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

- **Author:** [davila7](https://github.com/davila7)
- **Source:** [davila7/claude-with-skills](https://github.com/davila7/claude-with-skills)
- **License:** MIT
- **Homepage:** https://claude-with-skills.vercel.app

Install and usage instructions live in the source repository linked above.

## Pricing

- **Free** — Free

## Security capabilities

Automated source analysis of v0.1.0 — what this tool can access:

- **Network access:** no
- **Filesystem access:** no
- **Shell / process execution:** no
- **Environment & secrets:** no
- **Dynamic code execution:** no

*"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.*


## Versions

- **0.1.0** — security scan: passed — Imported from the upstream source.

## Links

- Listing page: https://agentstack.voostack.com/l/skill-davila7-claude-with-skills-summarize-changes
- Seller: https://agentstack.voostack.com/s/davila7
- Browse the marketplace: https://agentstack.voostack.com/browse

---
Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.