# 使用 Homebrew 安装 bazel-diff

查看 bazel-diff 的安装路径、可执行文件、元数据以及面向 AI 代理工作流的安全说明。

## 安装

```sh
sudo av install brew:bazel-diff
```

其他安装命令:

### macOS

- Homebrew (100%):

```sh
brew install bazel-diff
```

  证据: local Homebrew formula metadata

## 软件包事实

- **软件包键:** brew:bazel-diff
- **软件包管理器:** Homebrew
- **版本:** 39.0.0
- **来源摘要:** Performs Bazel Target Diffing between two revisions in Git
- **主页:** <https://github.com/Tinder/bazel-diff/>
- **仓库:** <https://github.com/Tinder/bazel-diff>
- **最后更新:** 2026-08-03T03:10:13Z
- **已生成:** 2026-08-03T19:37:03+00:00

## 可执行文件

- bazel-diff (别名)

## 安装行为

- Bottle: 不可用

## 版本和新鲜度

- 页面生成时间: 2026-08-03
- 管理器版本: 39.0.0
## 项目历史与用法

bazel-diff is Tinder's command-line utility for computing the Bazel targets affected between two Git revisions, letting CI systems build or test the precise changed target set instead of rerunning an entire monorepo.

### 项目历史

The upstream repository begins in September 2020 with a v1.0-era implementation of Bazel target diffing. Its README frames the project as an answer to a gap in Bazel itself, pointing to the long-running Bazel feature request for native target diffing and to a BazelCon talk as design inspiration.

The tool's implementation history is tied to very large Bazel graphs: the README calls out streamed protobuf query output, avoidance of long command-line query lists, support for renamed/deleted/modified files, and testing against projects with tens of thousands of targets.

### 采用历史

Adoption is strongest among Bazel monorepo users who need deterministic affected-target selection for pull-request CI. The Homebrew formula packages it as a standalone CLI, making it accessible outside the original Tinder repository.

Later README additions describe an experimental HTTP query service and cached hash generation, showing a shift from one-shot local comparison toward always-on CI infrastructure for large repositories.

### 使用方式

A typical workflow checks out a base revision, generates Bazel graph hashes, checks out a final revision, generates hashes again, and runs get-impacted-targets to emit the changed target labels as JSON.

The tool is used when correctness matters more than approximate file globs: its docs explicitly warn that incorrect affected-test selection erodes trust in green-to-red CI transitions.

### 为什么软件包爱好者会关心

For package-history nerds, bazel-diff is a good example of the post-Bazel ecosystem filling operational gaps around monorepo CI rather than replacing Bazel itself.

It also illustrates why Bazel-adjacent packages often matter more as automation primitives than as developer-facing build tools: the interesting artifact is the affected-target JSON consumed by CI.

### 时间线

- 2020-09: Upstream git history starts with the initial v1.0-era bazel-diff code.
- 2020s: Homebrew packages bazel-diff as a standalone CLI for macOS and Linux users.
- 2020s: README-documented features expand to build graph distance metrics and an experimental long-running query service.

### Related projects

- Bazel supplies the query and build graph model that bazel-diff analyzes.
- Aspect's impacted command is documented by bazel-diff as a Starlark/Aspect CLI way to expose a similar impacted-target workflow.
- Bazelisk is commonly paired with bazel-diff in CI so the Bazel binary used for graph hashing matches the workspace's pinned version.

### 来源

- Upstream git history for https://github.com/Tinder/bazel-diff starts in September 2020.
- <https://github.com/Tinder/bazel-diff#readme documents origins, use cases, impacted-target workflow, scale notes, and related Bazel feature request.>
- source_facts.package-manager-url records the Homebrew packaging surface.


## 安全说明

没有找到 bazel-diff 的匹配本地密钥处理 manifest。Nucleus 软件包元数据仍在此发布，以便未来覆盖拥有稳定的软件包 URL。



## Combined YAML source

View the package source record on GitHub. [combined/bazel-diff.yml](https://github.com/mxcl/pkgdb/blob/main/combined/bazel-diff.yml)


## 来源

- pkg.so package database
- Geiger risk classifier
- curated package history
- pkgdb category and tag curation
- cross-ecosystem install command graph
