Now liveThe Skillselion MCP - thousands of ranked skills, loaded into your agent mid-task. No install.Get it →
bytedance avatar

Byted Security Skillsscanner

  • 31 installs
  • 411 repo stars
  • Updated August 4, 2026
  • bytedance/agentkit-samples

byted-security-skillsscanner is a Claude skill that scans other skills for security risks by uploading them to Volcengine's scanning service.

About

A skill that audits other skills for security risks by packaging a target skill directory and uploading it to a Volcengine scanning service. A developer uses it to periodically scan workspace skills against a security policy, or to check a skill they are developing. It returns a JSON list of findings that is rendered as a risk report by severity.

  • Packages and uploads a skill directory to Volcengine to scan it for security risks
  • Audits other skills in the workspace for suspicious network calls or sensitive operations
  • Outputs a risk report by level (High/Medium/Low) with rule name and remediation

Byted Security Skillsscanner by the numbers

  • 31 all-time installs (skills.sh)
  • Ranked #1,486 of 2,203 Security skills by installs in the Skillselion catalog
  • Data as of Aug 5, 2026 (Skillselion catalog sync)
At a glance

byted-security-skillsscanner capabilities & compatibility

Capabilities
security audit
Use cases
security audit
From the docs

What byted-security-skillsscanner says it does

审计和扫描技能的安全性。当用户希望对工作区中的其他技能进行安全扫描时使用。
SKILL.md
必须确保目标技能包含 `SKILL.md` 文件,因为它是扫描的主要输入。
SKILL.md
npx skills add https://github.com/bytedance/agentkit-samples --skill byted-security-skillsscanner

Add your badge

Show developers this skill is listed on Skillselion. Paste this into your README.

Listed on Skillselion
Installs31
repo stars411
Last updatedAugust 4, 2026
Repositorybytedance/agentkit-samples

What it does

Package a target skill directory, upload it to the Volcengine scanner, and report security risks by severity.

Who is it for?

auditing workspace skills or your own skill for security risks before use

Skip if: scanning targets that have no SKILL.md file

When should I use this skill?

user asks to security-scan or audit a skill directory

What you get

The target skill is scanned and a risk report is produced by severity with per-finding remediation advice.

  • security scan report listing High/Medium risks with rule, file, detail, and remediation

By the numbers

  • 3 required environment variables
  • 3 risk levels (High/Medium/Low)

Files

SKILL.mdMarkdownGitHub ↗

Volcengine技能安全扫描器 (Volcengine Skills Scanner)

通过将技能目录打包并上传火山引擎安全扫描服务进行扫描,审计工作区中的其他技能是否存在潜在的安全风险。

何时使用

  • 审计:定期扫描所有技能以确保符合安全策略。
  • 开发:在开发过程中检查自己的技能。
  • 要求:必须确保目标技能包含 SKILL.md 文件,因为它是扫描的主要输入。

用法

使用 scripts/scan.py 脚本执行扫描。必须使用绝对路径,不要使用~,因为运行目录不是 skill 目录。

脚本会自动打包目录(如果提供的是目录)并上传,始终输出包含扫描结果的 JSON 数组。解析此JSON并以易读的格式(中文)向用户展示结果(风险等级、详细信息、建议)。

扫描技能(目录或压缩包)

脚本通过环境变量读取配置(推荐)

python3 ~/.openclaw/workspace/skills/byted-security-skillsscanner/scripts/scan.py --name "bad_skills1" --path "/root/.openclaw/workspace/skills/bad_skills1"

重要

  • 脚本路径必须是绝对路径
  • 目标路径也必须是绝对路径
  • 确保已设置必要的环境变量(VOLCENGINE_ACCESS_KEYVOLCENGINE_SECRET_KEYVOLCENGINE_REGION

报告格式

向用户展示结果时,必须使用以下格式(中文):

🛡️ 安全扫描报告:[SkillName]

扫描时间: [将 ScanEndTime 时间戳转换为可读日期格式] 整体状态: [✅ 通过 / ❌ 发现风险]

风险等级规则名称风险详情
[High/Medium/Low][RuleName][RiskDetail]

发现的风险列表: (仅列出 High 和 Medium 级别的风险)

1. [RuleName] (ID: [RuleID])

  • 等级: [RiskLevel]
  • 文件: [FileName]
  • 详情: [RiskDetail]
  • 建议: 请检查上述文件中的代码,移除可疑的网络请求或敏感操作。

---

环境变量配置

1. 获取火山引擎访问凭证:参考 用户指南 获取 AK/SK

2. 配置以下环境变量:

export VOLC_ACCESS_KEY="your-access-key"
export VOLC_SECRET_KEY="your-secret-key"
export VOLC_REGION="cn-north-1"  # 可选,默认 cn-north-1

Related skills

FAQ

What does the scanner need as input?

A skill directory or zip; the target must contain a SKILL.md file, since it is the main scan input.

How are results shown?

As a Chinese report grouping High and Medium risks with rule name, file, detail, and remediation.

Securityauditappsec

This week in AI coding

Five minutes, every Monday - the tools, releases and tactics for developers.

unsubscribe anytime.