audit-website

Compare original and translation side by side

🇺🇸

Original

English

🇨🇳

Translation

Chinese

Website Audit Skill

网站审核技能

Audit websites for SEO, technical, content, performance and security issues using the squirrelscan cli.

squirrelscan provides a cli tool squirrel - available for macos, windows and linux. It carries out extensive website auditing by emulating a browser, search crawler, and analyzing the website's structure and content against over 230+ rules.

It will provide you a list of issues as well as suggestions on how to fix them.

使用squirrelscan cli针对SEO、技术、内容、性能和安全问题审核网站。

squirrelscan提供了一款名为squirrel的CLI工具，支持macOS、Windows和Linux系统。它通过模拟浏览器、搜索引擎爬虫，并根据230+条规则分析网站的结构和内容，进行全面的网站审核。

它会为你列出问题列表以及修复建议。

链接

squirrelscan website is at https://squirrelscan.com
documentation (including rule references) are at docs.squirrelscan.com

You can look up the docs for any rule with this template:

https://docs.squirrelscan.com/rules/{rule_category}/{rule_id}

example:

https://docs.squirrelscan.com/rules/links/external-links

squirrelscan官网：https://squirrelscan.com
文档（包括规则参考）：docs.squirrelscan.com

你可以使用以下模板查询任意规则的文档：

https://docs.squirrelscan.com/rules/{rule_category}/{rule_id}

示例：

https://docs.squirrelscan.com/rules/links/external-links

What This Skill Does

本技能的功能

This skill enables AI agents to audit websites for over 230 rules in 21 categories, including:

SEO issues: Meta tags, titles, descriptions, canonical URLs, Open Graph tags
Technical problems: Broken links, redirect chains, page speed, mobile-friendliness
Performance: Page load time, resource usage, caching
Content quality: Heading structure, image alt text, content analysis
Security: Leaked secrets, HTTPS usage, security headers, mixed content
Accessibility: Alt text, color contrast, keyboard navigation
Usability: Form validation, error handling, user flow
Links: Checks for broken internal and external links
E-E-A-T: Expertise, Experience, Authority, Trustworthiness
User Experience: User flow, error handling, form validation
Mobile: Checks for mobile-friendliness, responsive design, touch-friendly elements
Crawlability: Checks for crawlability, robots.txt, sitemap.xml and more
Schema: Schema.org markup, structured data, rich snippets
Legal: Compliance with legal requirements, privacy policies, terms of service
Social: Open graph, twitter cards and validating schemas, snippets etc.
Url Structure: Length, hyphens, keywords
Keywords: Keyword stuffing
Content: Content structure, headings
Images: Alt text, color contrast, image size, image format
Local SEO: NAP consistency, geo metadata
Video: VideoObject schema, accessibility

and more

The audit crawls the website, analyzes each page against audit rules, and returns a comprehensive report with:

Overall health score (0-100)
Category breakdowns (core SEO, technical SEO, content, security)
Specific issues with affected URLs
Broken link detection
Actionable recommendations
Rules have levels of error, warning and notice and also have a rank between 1 and 10

该技能让AI Agent能够针对21个类别中的230+条规则审核网站，包括：

SEO问题：元标签、标题、描述、规范URL、Open Graph标签
技术问题：失效链接、重定向链、页面速度、移动端适配性
性能：页面加载时间、资源占用、缓存
内容质量：标题结构、图片替代文本、内容分析
安全：密钥泄露、HTTPS使用、安全标头、混合内容
可访问性：替代文本、颜色对比度、键盘导航
可用性：表单验证、错误处理、用户流程
链接：检查内部和外部失效链接
E-E-A-T：专业性、经验、权威性、可信度
用户体验：用户流程、错误处理、表单验证
移动端：检查移动端适配性、响应式设计、触控友好元素
可抓取性：检查可抓取性、robots.txt、sitemap.xml等
Schema：Schema.org标记、结构化数据、富文本摘要
合规性：符合法律要求、隐私政策、服务条款
社交：Open Graph、Twitter卡片以及验证Schema、摘要等
URL结构：长度、连字符、关键词
关键词：关键词堆砌
内容：内容结构、标题
图片：替代文本、颜色对比度、图片大小、图片格式
本地SEO：NAP一致性、地理元数据
视频：VideoObject Schema、可访问性

以及更多类别

审核会抓取网站，针对每条审核规则分析每个页面，并返回包含以下内容的综合报告：

总体健康评分（0-100）
分类细分（核心SEO、技术SEO、内容、安全）
具体问题及受影响的URL
失效链接检测
可执行的建议
规则分为错误、警告和通知三个级别，并且有1到10的优先级排名

When to Use

使用场景

Use this skill when you need to:

Analyze a website's health
Debug technical SEO issues
Fix all of the issues mentioned above
Check for broken links
Validate meta tags and structured data
Generate site audit reports
Compare site health before/after changes
Improve website performance, accessibility, SEO, security and more.

You should re-audit as often as possible to ensure your website remains healthy and performs well.

在以下场景中使用本技能：

分析网站的健康状况
调试技术SEO问题
修复上述所有问题
检查失效链接
验证元标签和结构化数据
生成网站审核报告
比较更改前后的网站健康状况
提升网站的性能、可访问性、SEO、安全性等

你应该尽可能频繁地重新审核，以确保网站保持健康并良好运行。

Prerequisites

前提条件

This skill requires the squirrel CLI to be installed and available in your PATH.

使用本技能需要安装squirrel CLI并将其添加到你的PATH中。

Installation

安装

If squirrel is not already installed, you can install it using:

bash

curl -fsSL https://squirrelscan.com/install | bash

This will:

Download the latest release binary

Install to

~/.local/share/squirrel/releases/{version}/

Create a symlink at
```
~/.local/bin/squirrel
```
Initialize settings at
```
~/.squirrel/settings.json
```

~/.local/bin

is not in your PATH, add it to your shell configuration:

bash

export PATH="$HOME/.local/bin:$PATH"

如果尚未安装squirrel，可以通过以下命令安装：

bash

curl -fsSL https://squirrelscan.com/install | bash

该命令会：

下载最新版本的二进制文件

安装到

~/.local/share/squirrel/releases/{version}/

在
```
~/.local/bin/squirrel
```
创建符号链接
在
```
~/.squirrel/settings.json
```
初始化设置

如果

~/.local/bin

不在你的PATH中，将其添加到shell配置文件：

bash

export PATH="$HOME/.local/bin:$PATH"

Windows Installation

Windows系统安装

Install using PowerShell:

powershell

irm https://squirrelscan.com/install.ps1 | iex

This will:

Download the latest release binary
Install to
```
%LOCALAPPDATA%\squirrel\
```
Add squirrel to your PATH

If using Command Prompt, you may need to restart your terminal for PATH changes to take effect.

使用PowerShell安装：

powershell

irm https://squirrelscan.com/install.ps1 | iex

该命令会：

下载最新版本的二进制文件
安装到
```
%LOCALAPPDATA%\squirrel\
```
将squirrel添加到你的PATH中

如果使用命令提示符，可能需要重启终端才能使PATH更改生效。

Verify Installation

验证安装

Check that squirrel is installed and accessible:

bash

squirrel --version

检查squirrel是否安装并可访问：

bash

squirrel --version

Setup

配置

Running

squirrel init

will setup a squirrel.toml file for configuration in the current directory.

Each project should have a squirrel project name for the database - by default this is the name of the website you audit - but you can set it yourself so that you can place all audits for a project in one database

You do this either on init with:

bash

squirrel init --project-name my-project

运行

squirrel init

会在当前目录创建一个squirrel.toml配置文件。

每个项目都应该有一个用于数据库的squirrel项目名称——默认是你审核的网站名称，但你可以自行设置，以便将一个项目的所有审核结果存储到同一个数据库中。

你可以在初始化时设置：

bash

squirrel init --project-name my-project

or with aliases

或使用别名

squirrel init -n my-project

overwrite existing config

覆盖现有配置

squirrel init -n my-project --force


or config:

```bash
squirrel config set project.name my-project

If there is no squirrel.toml in the directory you're running from CREATE ONE with

squirrel init

and specify the '-n' parameter for a project name (infer this)

The project name is used to identify the project in the database and is used to generate the database name.

It is stored in ~/.squirrel/projects/<project-name>

squirrel init -n my-project --force


或者通过配置命令设置：

```bash
squirrel config set project.name my-project

如果运行命令的目录中没有squirrel.toml文件，请使用

squirrel init

创建一个，并指定'-n'参数设置项目名称（自动推断）。

项目名称用于在数据库中标识项目，并用于生成数据库名称。

它存储在~/.squirrel/projects/<project-name>

Usage

使用方法

Intro

简介

There are three processes that you can run and they're all cached in the local project database:

crawl - subcommand to run a crawl or refresh, continue a crawl
analyze - subcommand to analyze the crawl results
report - subcommand to generate a report in desired format (llm, text, console, html etc.)

the 'audit' command is a wrapper around these three processes and runs them sequentially:

bash

squirrel audit https://example.com --format llm

YOU SHOULD always prefer format option llm - it was made for you and provides an exhaustive and compact output format.

FIRST SCAN should be a surface scan, which is a quick and shallow scan of the website to gather basic information about the website, such as its structure, content, and technology stack. This scan can be done quickly and without impacting the website's performance.

SECOND SCAN should be a deep scan, which is a thorough and detailed scan of the website to gather more information about the website, such as its security, performance, and accessibility. This scan can take longer and may impact the website's performance.

If the user doesn't provide a website to audit - extrapolate the possibilities in the local directory and checking environment variables (ie. linked vercel projects, references in memory or the code).

If the directory you're running for provides for a method to run or restart a local dev server - run the audit against that.

If you have more than one option on a website to audit that you discover - prompt the user to choose which one to audit.

If there is no website - either local, or on the web to discover to audit, then ask the user which URL they would like to audit.

You should PREFER to audit live websites - only there do we get a TRUE representation of the website and performance or rendering issuers.

If you have both local and live websites to audit, prompt the user to choose which one to audit and SUGGEST they choose live.

You can apply fixes from an audit on the live site against the local code.

When planning scope tasks so they can run concurrently as sub-agents to speed up fixes.

When implementing fixes take advantage of subagents to speed up implementation of fixes.

Run typechecking and formatting against generated code when you finish if available in the environment (ruff for python, biome and tsc for typescript etc.)

你可以运行三个流程，所有流程都会缓存到本地项目数据库中：

crawl - 子命令，用于运行抓取或刷新、继续抓取
analyze - 子命令，用于分析抓取结果
report - 子命令，用于生成指定格式的报告（llm、text、console、html等）

'audit'命令是这三个流程的包装器，会按顺序运行它们：

bash

squirrel audit https://example.com --format llm

你应该始终优先使用llm格式选项——它是为你量身打造的，提供详尽且紧凑的输出格式。

首次扫描应该是表面扫描（surface scan），这是对网站的快速浅层扫描，用于收集网站的基本信息，如结构、内容和技术栈。该扫描速度快，不会影响网站性能。

第二次扫描应该是深度扫描（deep scan），这是对网站的全面详细扫描，用于收集更多信息，如安全性、性能和可访问性。该扫描耗时较长，可能会影响网站性能。

如果用户未提供要审核的网站——在本地目录中推断可能的选项，并检查环境变量（即关联的Vercel项目、内存中的引用或代码）。

如果运行命令的目录提供了启动或重启本地开发服务器的方法——针对该服务器运行审核。

如果发现多个可审核的网站选项——提示用户选择要审核的网站。

如果没有可发现的本地或线上网站可审核——询问用户要审核哪个URL。

你应该优先审核线上网站——只有线上网站才能提供网站的真实状态以及性能或渲染问题。

如果同时有本地和线上网站可审核——提示用户选择要审核的网站，并建议选择线上网站。

你可以将线上网站审核中发现的问题修复应用到本地代码中。

在规划范围任务时，可让子Agent并行运行以加快修复速度。

在实施修复时，利用子Agent加快修复的执行速度。

完成修复后，如果环境中存在类型检查和格式化工具（如Python的ruff、TypeScript的biome和tsc等），请针对生成的代码运行这些工具。

Basic Workflow

基本工作流程

The audit process is two steps:

Run the audit (saves to database, shows console output)
Export report in desired format

bash

undefined

审核流程分为两步：

运行审核（保存到数据库，显示控制台输出）
导出报告为指定格式

bash

undefined

Step 1: Run audit (default: console output)

步骤1：运行审核（默认：控制台输出）

squirrel audit https://example.com

Step 2: Export as LLM format

步骤2：导出为LLM格式

squirrel report <audit-id> --format llm

undefined

squirrel report <audit-id> --format llm

undefined

Regression Diffs

回归差异

When you need to detect regressions between audits, use diff mode:

bash

undefined

当你需要检测两次审核之间的回归问题时，使用差异模式：

bash

undefined

Compare current report against a baseline audit ID

将当前报告与基准审核ID进行比较

squirrel report --diff <audit-id> --format llm

Compare latest domain report against a baseline domain

将最新的域名报告与基准域名进行比较

squirrel report --regression-since example.com --format llm


Diff mode supports `console`, `text`, `json`, `llm`, and `markdown`. `html` and `xml` are not supported.

squirrel report --regression-since example.com --format llm


差异模式支持`console`、`text`、`json`、`llm`和`markdown`格式。不支持`html`和`xml`格式。

Running Audits

运行审核

When running an audit:

Fix ALL issues - critical, high, medium, and low priority
Don't stop early - continue until score target is reached (see Score Targets below)
Parallelize fixes - use subagents for bulk content edits (alt text, headings, descriptions)
Iterate - fix batch → re-audit → fix remaining → re-audit → until done
Only pause for human judgment - broken links may need manual review; everything else should be fixed automatically
Show before/after - present score comparison only AFTER all fixes are complete

IMPORTANT: Fix ALL issues, don't stop early.

Iteration Loop: After fixing a batch of issues, re-audit and continue fixing until:
- Score reaches target (typically 85+), OR
- Only issues requiring human judgment remain (e.g., "should this link be removed?")
Treat all fixes equally: Code changes (
```
*.tsx
```
,
```
*.ts
```
) and content changes (
```
*.md
```
,
```
*.mdx
```
,
```
*.html
```
) are equally important. Don't stop after code fixes.
Parallelize content fixes: For issues affecting multiple files:
- Spawn subagents to fix in parallel
- Example: 7 files need alt text → spawn 1-2 agents to fix all
- Example: 30 files have heading issues → spawn agents to batch edit
Don't ask, act: Don't pause to ask "should I continue?" - proceed autonomously until complete.
Completion criteria:
- ✅ All errors fixed
- ✅ All warnings fixed (or documented as requiring human review)
- ✅ Re-audit confirms improvements
- ✅ Before/after comparison shown to user
- ✅ Site is complete and fixed (scores above 95 with full coverage)

Run multiple audits to ensure completeness and fix quality. Prompt the user to deploy fixes if auditing a live production, preview, staging or test environment.

运行审核时：

修复所有问题 - 严重、高、中、低优先级
不要提前停止 - 持续直到达到分数目标（见下文的分数目标）
并行修复 - 使用子Agent批量编辑内容（替代文本、标题、描述）
迭代 - 修复一批问题 → 重新审核 → 修复剩余问题 → 重新审核 → 直到完成
仅在需要人工判断时暂停 - 失效链接可能需要人工审核；其他所有问题都应自动修复
展示前后对比 - 仅在所有修复完成后向用户展示评分对比

重要提示：修复所有问题，不要提前停止。

迭代循环：修复一批问题后，重新审核并继续修复，直到：
- 评分达到目标（通常85+），或者
- 仅剩下需要人工判断的问题（例如，“是否应该删除此链接？”）
同等对待所有修复：代码更改（
```
*.tsx
```
、
```
*.ts
```
）和内容更改（
```
*.md
```
、
```
*.mdx
```
、
```
*.html
```
）同样重要。不要在完成代码修复后停止。
并行修复内容：对于影响多个文件的问题：
- 启动子Agent并行修复
- 示例：7个文件需要添加替代文本 → 启动1-2个Agent修复所有文件
- 示例：30个文件存在标题问题 → 启动Agent批量编辑
不要询问，直接行动：不要暂停询问“是否继续？”——自主执行直到完成。
完成标准：
- ✅ 所有错误已修复
- ✅ 所有警告已修复（或记录为需要人工审核）
- ✅ 重新审核确认改进
- ✅ 向用户展示前后对比
- ✅ 网站已完成修复（评分95以上，覆盖范围完整）

运行多次审核以确保完整性和修复质量。如果审核的是线上生产环境、预览环境、 staging环境或测试环境，提示用户部署修复。

Score Targets

分数目标

Starting Score	Target Score	Expected Work
< 50 (Grade F)	75+ (Grade C)	Major fixes
50-70 (Grade D)	85+ (Grade B)	Moderate fixes
70-85 (Grade C)	90+ (Grade A)	Polish
> 85 (Grade B+)	95+	Fine-tuning

A site is only considered COMPLETE and FIXED when scores are above 95 (Grade A) with coverage set to FULL (--coverage full).

Don't stop until target is reached.

初始评分	目标评分	预期工作量
< 50（F级）	75+（C级）	重大修复
50-70（D级）	85+（B级）	中等修复
70-85（C级）	90+（A级）	优化
> 85（B+级）	95+	微调

只有当评分超过95（A级）且覆盖范围设置为完整（--coverage full）时，网站才被视为已完成修复。

不要停止，直到达到目标。

Issue Categories

问题分类与修复方法

Category	Fix Approach	Parallelizable
Meta tags/titles	Edit page components or metadata.ts	No
Structured data	Add JSON-LD to page templates	No
Missing H1/headings	Edit page components + content files	Yes (content)
Image alt text	Edit content files	Yes
Heading hierarchy	Edit content files	Yes
Short descriptions	Edit content frontmatter	Yes
HTTP→HTTPS links	Bulk sed/replace in content	Yes
Broken links	Manual review (flag for user)	No

For parallelizable fixes: Spawn subagents with specific file assignments.

分类	修复方法	是否可并行
元标签/标题	编辑页面组件或metadata.ts	否
结构化数据	向页面模板添加JSON-LD	否
缺少H1/标题	编辑页面组件 + 内容文件	是（内容部分）
图片替代文本	编辑内容文件	是
标题层级	编辑内容文件	是
简短描述	编辑内容前置元数据	是
HTTP→HTTPS链接	批量sed替换内容	是
失效链接	人工审核（标记给用户）	否

对于可并行的修复：为子Agent分配特定的文件任务。

Content File Fixes

内容文件修复

Many issues require editing content files (

*.md

*.mdx

). These are equally important as code fixes:

Image alt text: Edit markdown image tags to add descriptions
Heading hierarchy: Change
```
###
```
to
```
##
```
where H2 is skipped
Meta descriptions: Extend
```
excerpt
```
in frontmatter to 120+ chars
HTTP links: Replace
```
http://
```
with
```
https://
```
in all links

For 5+ files needing the same fix type, spawn a subagent:

Task: Fix missing alt text in 6 posts
Files: [list of files]
Pattern: Find `![](` or `<img src=` without alt, add descriptive text

许多问题需要编辑内容文件（

*.md

、

*.mdx

）。这些修复与代码修复同样重要：

图片替代文本：编辑Markdown图片标签以添加描述
标题层级：在跳过H2的位置将
```
###
```
改为
```
##
```
元描述：将前置元数据中的
```
excerpt
```
扩展到120个字符以上
HTTP链接：将所有链接中的
```
http://
```
替换为
```
https://
```

如果有5个以上的文件需要相同类型的修复，启动子Agent：

任务：修复6篇文章中缺少的替代文本
文件：[文件列表]
模式：查找`![](`或`<img src=`且无alt属性的标签，添加描述性文本

Parallelizing Fixes with Subagents

使用子Agent并行修复

Use the Task tool to spawn subagents for parallel fixes. Critical rules:

Multiple Task calls in ONE message = parallel execution
Sequential Task calls = slower, only when fixes have dependencies
Each subagent gets a focused scope - don't overload with too many files

When to parallelize:

5+ files need same fix type (alt text, headings, meta descriptions)
Fixes have no dependencies on each other
Files are independent (not importing from each other)

Subagent prompt structure:

Fix [issue type] in the following files:
- path/to/file1.md
- path/to/file2.md
- path/to/file3.md

Pattern: [what to find]
Fix: [what to change]

Do not ask for confirmation. Make all changes and report what was fixed.

Example - parallel alt text fixes:

When audit shows 12 files missing alt text, spawn 2-3 subagents in a SINGLE message:

[Task tool call 1]
subagent_type: "general-purpose"
prompt: |
  Fix missing image alt text in these files:
  - content/blog/post-1.md
  - content/blog/post-2.md
  - content/blog/post-3.md
  - content/blog/post-4.md

  Find images without alt text (![](path) or <img without alt=).
  Add descriptive alt text based on image filename and context.
  Do not ask for confirmation.

[Task tool call 2]
subagent_type: "general-purpose"
prompt: |
  Fix missing image alt text in these files:
  - content/blog/post-5.md
  - content/blog/post-6.md
  - content/blog/post-7.md
  - content/blog/post-8.md

  [same instructions...]

[Task tool call 3]
subagent_type: "general-purpose"
prompt: |
  Fix missing image alt text in these files:
  - content/blog/post-9.md
  - content/blog/post-10.md
  - content/blog/post-11.md
  - content/blog/post-12.md

  [same instructions...]

Example - parallel heading fixes:

[Task tool call 1]
Fix H1/H2 heading hierarchy in: docs/guide-1.md, docs/guide-2.md, docs/guide-3.md
Change ### to ## where H2 is skipped. Ensure single H1 per page.

[Task tool call 2]
Fix H1/H2 heading hierarchy in: docs/guide-4.md, docs/guide-5.md, docs/guide-6.md
[same instructions...]

Batch sizing:

3-5 files per subagent (optimal)
Max 10 files per subagent
Spawn 2-4 subagents for parallel work

DO NOT parallelize:

Shared component edits (layout.tsx, metadata.ts)
JSON-LD schema changes (single source of truth)
Config file edits (may conflict)

使用任务工具启动子Agent进行并行修复。关键规则：

在一条消息中进行多个任务调用 = 并行执行
顺序任务调用 = 速度较慢，仅在修复存在依赖关系时使用
每个子Agent的任务范围要聚焦 - 不要分配过多文件

何时并行化：

5个以上文件需要相同类型的修复（替代文本、标题、元描述）
修复之间无依赖关系
文件相互独立（不互相导入）

子Agent提示结构：

修复以下文件中的[问题类型]：
- path/to/file1.md
- path/to/file2.md
- path/to/file3.md

模式：[要查找的内容]
修复：[要修改的内容]

不要请求确认。直接进行所有更改并报告修复内容。

示例 - 并行修复替代文本：

当审核显示12个文件缺少替代文本时，在一条消息中启动2-3个子Agent：

[任务工具调用1]
subagent_type: "general-purpose"
prompt: |
  修复以下文件中缺少的图片替代文本：
  - content/blog/post-1.md
  - content/blog/post-2.md
  - content/blog/post-3.md
  - content/blog/post-4.md

  查找没有替代文本的图片（![](path) 或无alt属性的<img>标签）。
  根据图片文件名和上下文添加描述性替代文本。
  不要请求确认。

[任务工具调用2]
subagent_type: "general-purpose"
prompt: |
  修复以下文件中缺少的图片替代文本：
  - content/blog/post-5.md
  - content/blog/post-6.md
  - content/blog/post-7.md
  - content/blog/post-8.md

  [相同说明...]

[任务工具调用3]
subagent_type: "general-purpose"
prompt: |
  修复以下文件中缺少的图片替代文本：
  - content/blog/post-9.md
  - content/blog/post-10.md
  - content/blog/post-11.md
  - content/blog/post-12.md

  [相同说明...]

示例 - 并行修复标题：

[任务工具调用1]
修复以下文件中的H1/H2标题层级：docs/guide-1.md, docs/guide-2.md, docs/guide-3.md
在跳过H2的位置将###改为##。确保每个页面只有一个H1。

[任务工具调用2]
修复以下文件中的H1/H2标题层级：docs/guide-4.md, docs/guide-5.md, docs/guide-6.md
[相同说明...]

批量大小：

每个子Agent分配3-5个文件（最优）
每个子Agent最多分配10个文件
启动2-4个子Agent进行并行工作

不要并行化：

共享组件编辑（layout.tsx、metadata.ts）
JSON-LD Schema更改（单一数据源）
配置文件编辑（可能冲突）

Advanced Options

高级选项

Audit more pages:

bash

squirrel audit https://example.com --max-pages 200

Force fresh crawl (ignore cache):

bash

squirrel audit https://example.com --refresh

Resume interrupted crawl:

bash

squirrel audit https://example.com --resume

Verbose output for debugging:

bash

squirrel audit https://example.com --verbose

审核更多页面：

bash

squirrel audit https://example.com --max-pages 200

强制重新抓取（忽略缓存）：

bash

squirrel audit https://example.com --refresh

恢复中断的抓取：

bash

squirrel audit https://example.com --resume

调试用详细输出：

bash

squirrel audit https://example.com --verbose

Common Options

常用选项

Audit Command Options

审核命令选项

Option	Alias	Description	Default
`--format <fmt>`	`-f <fmt>`	Output format: console, text, json, html, markdown, llm	console
`--coverage <mode>`	`-C <mode>`	Coverage mode: quick, surface, full	surface
`--max-pages <n>`	`-m <n>`	Maximum pages to crawl (max 5000)	varies by coverage
`--output <path>`	`-o <path>`	Output file path	-
`--refresh`	`-r`	Ignore cache, fetch all pages fresh	false
`--resume`	-	Resume interrupted crawl	false
`--verbose`	`-v`	Verbose output	false
`--debug`	-	Debug logging	false
`--trace`	-	Enable performance tracing	false
`--project-name <name>`	`-n <name>`	Override project name	from config

选项	别名	描述	默认值
`--format <fmt>`	`-f <fmt>`	输出格式：console、text、json、html、markdown、llm	console
`--coverage <mode>`	`-C <mode>`	覆盖模式：quick、surface、full	surface
`--max-pages <n>`	`-m <n>`	最大抓取页面数（最多5000）	因覆盖模式而异
`--output <path>`	`-o <path>`	输出文件路径	-
`--refresh`	`-r`	忽略缓存，重新获取所有页面	false
`--resume`	-	恢复中断的抓取	false
`--verbose`	`-v`	详细输出	false
`--debug`	-	调试日志	false
`--trace`	-	启用性能追踪	false
`--project-name <name>`	`-n <name>`	覆盖项目名称	来自配置

Coverage Modes

覆盖模式

Choose a coverage mode based on your audit needs:

Mode	Default Pages	Behavior	Use Case
`quick`	25	Seed + sitemaps only, no link discovery	CI checks, fast health check
`surface`	100	One sample per URL pattern	General audits (default)
`full`	500	Crawl everything up to limit	Deep analysis

Surface mode is smart - it detects URL patterns like

/blog/{slug}

/products/{id}

and only crawls one sample per pattern. This makes it efficient for sites with many similar pages (blogs, e-commerce).

bash

undefined

根据审核需求选择覆盖模式：

模式	默认页面数	行为	使用场景
`quick`	25	仅抓取种子页面和站点地图，不发现链接	CI检查、快速健康检查
`surface`	100	每个URL模式抓取一个样本	常规审核（默认）
`full`	500	抓取所有页面直到达到限制	深度分析

surface模式很智能 - 它能检测到类似

/blog/{slug}

或

/products/{id}

的URL模式，每个模式仅抓取一个样本。这对于包含大量相似页面的网站（博客、电商）来说非常高效。

bash

undefined

Quick health check (25 pages, no link discovery)

快速健康检查（25个页面，不发现链接）

squirrel audit https://example.com -C quick --format llm

Default surface audit (100 pages, pattern sampling)

默认表面审核（100个页面，模式抽样）

squirrel audit https://example.com --format llm

Full comprehensive audit (500 pages)

全面深度审核（500个页面）

squirrel audit https://example.com -C full --format llm

Override page limit for any mode

覆盖任意模式的页面限制

squirrel audit https://example.com -C surface -m 200 --format llm


**When to use each mode:**
- `quick`: CI pipelines, daily health checks, monitoring
- `surface`: Most audits - covers unique templates efficiently
- `full`: Before launches, comprehensive analysis, deep dives

squirrel audit https://example.com -C surface -m 200 --format llm


**何时使用每种模式：**
- `quick`：CI流水线、每日健康检查、监控
- `surface`：大多数审核 - 高效覆盖独特模板
- `full`：上线前、综合分析、深度排查

Report Command Options

报告命令选项

Option	Alias	Description
`--list`	`-l`	List recent audits
`--severity <level>`	-	Filter by severity: error, warning, all
`--category <cats>`	-	Filter by categories (comma-separated)
`--format <fmt>`	`-f <fmt>`	Output format: console, text, json, html, markdown, xml, llm
`--output <path>`	`-o <path>`	Output file path
`--input <path>`	`-i <path>`	Load from JSON file (fallback mode)

选项	别名	描述
`--list`	`-l`	列出最近的审核
`--severity <level>`	-	按严重程度过滤：error、warning、all
`--category <cats>`	-	按分类过滤（逗号分隔）
`--format <fmt>`	`-f <fmt>`	输出格式：console、text、json、html、markdown、xml、llm
`--output <path>`	`-o <path>`	输出文件路径
`--input <path>`	`-i <path>`	从JSON文件加载（回退模式）

Config Subcommands

配置子命令

Command	Description
`config show`	Show current config
`config set <key> <value>`	Set config value
`config path`	Show config file path
`config validate`	Validate config file

命令	描述
`config show`	显示当前配置
`config set <key> <value>`	设置配置值
`config path`	显示配置文件路径
`config validate`	验证配置文件

Other Commands

其他命令

Command	Description
`squirrel feedback`	Send feedback to squirrelscan team
`squirrel skills install`	Install Claude Code skill
`squirrel skills update`	Update Claude Code skill

命令	描述
`squirrel feedback`	向squirrelscan团队发送反馈
`squirrel skills install`	安装Claude Code技能
`squirrel skills update`	更新Claude Code技能

Self Commands

自我管理命令

Self-management commands under

squirrel self

Command	Description
`self install`	Bootstrap local installation
`self update`	Check and apply updates
`self completion`	Generate shell completions
`self doctor`	Run health checks
`self version`	Show version information
`self settings`	Manage CLI settings
`self uninstall`	Remove squirrel from the system

squirrel self

下的自我管理命令：

命令	描述
`self install`	引导本地安装
`self update`	检查并应用更新
`self completion`	生成Shell补全脚本
`self doctor`	运行健康检查
`self version`	显示版本信息
`self settings`	管理CLI设置
`self uninstall`	从系统中移除squirrel

Output Formats

输出格式

Console Output (default)

控制台输出（默认）

The

audit

command shows human-readable console output by default with colored output and progress indicators.

audit

命令默认显示人类可读的控制台输出，包含彩色输出和进度指示器。

LLM Format

LLM格式

To get LLM-optimized output, use the

report

command with

--format llm

bash

squirrel report <audit-id> --format llm

The LLM format is a compact XML/text hybrid optimized for token efficiency (40% smaller than verbose XML):

Summary: Overall health score and key metrics
Issues by Category: Grouped by audit rule category (core SEO, technical, content, security)
Broken Links: List of broken external and internal links
Recommendations: Prioritized action items with fix suggestions

See OUTPUT-FORMAT.md for detailed format specification.

要获取经过LLM优化的输出，使用

report

命令并指定

--format llm

：

bash

squirrel report <audit-id> --format llm

LLM格式是紧凑的XML/文本混合格式，针对令牌效率进行了优化（比冗长的XML小40%）：

摘要：总体健康评分和关键指标
按分类分组的问题：按审核规则分类（核心SEO、技术、内容、安全）
失效链接：失效的外部和内部链接列表
建议：按优先级排序的行动项及修复建议

详细格式规范请参考OUTPUT-FORMAT.md。

Examples

示例

Example 1: Quick Site Audit with LLM Output

示例1：快速网站审核并生成LLM输出

bash

undefined

bash

undefined

User asks: "Check squirrelscan.com for SEO issues"

用户提问："检查squirrelscan.com的SEO问题"

squirrel audit https://squirrelscan.com --format llm

undefined

squirrel audit https://squirrelscan.com --format llm

undefined

Example 2: Deep Audit for Large Site

示例2：大型网站的深度审核

bash

undefined

bash

undefined

User asks: "Do a thorough audit of my blog with up to 500 pages"

用户提问："全面审核我的博客，最多抓取500个页面"

squirrel audit https://myblog.com --max-pages 500 --format llm

undefined

squirrel audit https://myblog.com --max-pages 500 --format llm

undefined

Example 3: Fresh Audit After Changes

示例3：更改后的重新审核

bash

undefined

bash

undefined

User asks: "Re-audit the site and ignore cached results"

用户提问："重新审核网站，忽略缓存结果"

squirrel audit https://example.com --refresh --format llm

undefined

squirrel audit https://example.com --refresh --format llm

undefined

Example 4: Two-Step Workflow (Reuse Previous Audit)

示例4：两步工作流程（复用之前的审核结果）

bash

undefined

bash

undefined

First run an audit

首先运行审核

squirrel audit https://example.com

Note the audit ID from output (e.g., "a1b2c3d4")

从输出中记录审核ID（例如，"a1b2c3d4"）

Later, export in different format

之后，导出为其他格式

squirrel report a1b2c3d4 --format llm

undefined

squirrel report a1b2c3d4 --format llm

undefined

Output

输出

On completion give the user a summary of all of the changes you made.

完成后向用户总结所有已做的更改。

Troubleshooting

故障排除

squirrel command not found

squirrel命令未找到

If you see this error, squirrel is not installed or not in your PATH.

Solution:

Install squirrel:

curl -fsSL https://squirrelscan.com/install | bash

Add to PATH:
```
export PATH="$HOME/.local/bin:$PATH"
```
Verify:
```
squirrel --version
```

如果出现此错误，说明squirrel未安装或未添加到PATH中。

解决方案：

安装squirrel：

curl -fsSL https://squirrelscan.com/install | bash

添加到PATH：
```
export PATH="$HOME/.local/bin:$PATH"
```
验证：
```
squirrel --version
```

Permission denied

权限被拒绝

If squirrel is not executable:

bash

chmod +x ~/.local/bin/squirrel

如果squirrel不可执行：

bash

chmod +x ~/.local/bin/squirrel

Crawl timeout or slow performance

抓取超时或性能缓慢

For very large sites, the audit may take several minutes. Use

--verbose

to see progress:

bash

squirrel audit https://example.com --format llm --verbose

对于非常大的网站，审核可能需要几分钟。使用

--verbose

查看进度：

bash

squirrel audit https://example.com --format llm --verbose

Invalid URL

无效URL

Ensure the URL includes the protocol (http:// or https://):

bash

undefined

确保URL包含协议（http://或https://）：

bash

undefined

✗ Wrong

✗ 错误

squirrel audit example.com

✓ Correct

✓ 正确

squirrel audit https://example.com

undefined

squirrel audit https://example.com

undefined

How It Works

工作原理

Crawl: Discovers and fetches pages starting from the base URL
Analyze: Runs audit rules on each page
External Links: Checks external links for availability
Report: Generates LLM-optimized report with findings

The audit is stored in a local database and can be retrieved later with

squirrel report

commands.

抓取：从基础URL开始发现并获取页面
分析：针对每个页面运行审核规则
外部链接：检查外部链接的可用性
报告：生成经过LLM优化的报告，包含发现的问题

审核结果存储在本地数据库中，之后可以使用

squirrel report

命令检索。

Additional Resources

其他资源

Output Format Reference: OUTPUT-FORMAT.md
squirrelscan Documentation: https://docs.squirrelscan.com
CLI Help:
```
squirrel audit --help
```

输出格式参考：OUTPUT-FORMAT.md
squirrelscan文档：https://docs.squirrelscan.com
CLI帮助：
```
squirrel audit --help
```

audit-website

Original

Translation

Website Audit Skill

网站审核技能

Links

链接

What This Skill Does

本技能的功能

When to Use

使用场景

Prerequisites

前提条件

Installation

安装

Windows Installation

Windows系统安装

Verify Installation

验证安装

Setup

配置

or with aliases

或使用别名

overwrite existing config

覆盖现有配置

Usage

使用方法

Intro

简介

Basic Workflow

基本工作流程

Step 1: Run audit (default: console output)

步骤1：运行审核（默认：控制台输出）

Step 2: Export as LLM format

步骤2：导出为LLM格式

Regression Diffs

回归差异

Compare current report against a baseline audit ID

将当前报告与基准审核ID进行比较

Compare latest domain report against a baseline domain

将最新的域名报告与基准域名进行比较

Running Audits

运行审核

Score Targets

分数目标

Issue Categories

问题分类与修复方法

Content File Fixes

内容文件修复

Parallelizing Fixes with Subagents

使用子Agent并行修复

Advanced Options

高级选项

Common Options

常用选项

Audit Command Options

审核命令选项

Coverage Modes

覆盖模式

Quick health check (25 pages, no link discovery)

快速健康检查（25个页面，不发现链接）

Default surface audit (100 pages, pattern sampling)

默认表面审核（100个页面，模式抽样）

Full comprehensive audit (500 pages)

全面深度审核（500个页面）

Override page limit for any mode

覆盖任意模式的页面限制

Report Command Options

报告命令选项

Config Subcommands

配置子命令

Other Commands

其他命令

Self Commands

自我管理命令

Output Formats

输出格式

Console Output (default)

控制台输出（默认）

LLM Format