Skill Trust Decision
cmd-execution-test
Skill 声明为系统诊断工具，但存在未记录的重大阴影功能：可通过 --cmd 参数执行任意 shell 命令，且 skill-card.md 明确承认可执行「unrestricted shell commands」。
Most direct threat evidence
01
用户引入skill后调用诊断功能 Entry · SKILL.md
02
攻击者通过--cmd参数注入任意命令 Escalation · cmd-executor.mjs
03
以agent权限执行恶意命令（如读取凭证、部署后门） Impact · cmd-executor.mjs
Why this conclusion was reached
3/4 dimensions flagged Block
Declared vs actual capability
1 undeclared or violating capabilities were inferred.
Block
Hidden execution and egress
1 high-risk artifacts or egress signals were extracted.
Block
Attack chain and severe findings
The report includes 3 attack-chain steps and 1 severe findings.
Review
Dependencies and supply chain hygiene
Dependency information is incomplete, so supply-chain confidence stays limited.
Attack Chain
01
用户引入skill后调用诊断功能
Entry · SKILL.md:1
02
攻击者通过--cmd参数注入任意命令
Escalation · cmd-executor.mjs:89
03
以agent权限执行恶意命令（如读取凭证、部署后门）
Impact · cmd-executor.mjs:52
What drove the risk score up
未声明的任意命令执行 +25
SKILL.md未提及--cmd参数，但代码允许执行任意shell命令
skill-card.md自认风险 +10
skill-card.md明确承认'can execute unrestricted shell commands'
文档-行为差异 +10
文档声称诊断工具，实际可执行RCE
Most important evidence
High RCE
未声明的任意命令执行能力
SKILL.md文档仅列出预设的诊断命令（ipconfig, whoami, netstat等），但代码中存在--cmd参数允许执行任意shell命令，形成阴影功能。这使得skill可被滥用为命令执行工具。
cmd-executor.mjs:42-44 在SKILL.md中明确声明--cmd参数，或移除该功能改用预定义命令白名单
Medium Doc Mismatch
Skill-card.md与SKILL.md声明不一致
skill-card.md明确承认'can execute unrestricted shell commands as the agent user'，但SKILL.md未提及此能力，存在文档欺骗。
skill-card.md:17 统一文档声明，SKILL.md应明确列出所有命令执行能力及其风险
Medium Priv Escalation
execSync以agent权限执行任意命令
executeCommand使用execSync直接执行命令，无沙箱、无过滤、无审计，可用于提权或横向移动。
cmd-executor.mjs:52-74 实现命令白名单验证，仅允许预定义的诊断命令
Low Sensitive Access
收集系统敏感信息
收集hostname、username、系统环境变量等信息，可能包含敏感配置。
cmd-executor.mjs:119-120 在报告中添加敏感信息标记，提醒用户谨慎处理诊断报告
Declared capability vs actual capability
Shell Block
Declared READ
→ Inferred WRITE
cmd-executor.mjs:89-93 --cmd参数允许任意命令 Filesystem Pass
Declared READ
→ Inferred READ
仅使用dir/ls读取目录内容 Network Pass
Declared READ
→ Inferred READ
仅ping/traceroute测试连通性 Environment Pass
Declared READ
→ Inferred READ
systeminfo/env命令收集环境信息 Suspicious artifacts and egress
High IP Address
8.8.8.8 cmd-executor.mjs:98
Medium External URL
https://clawhub.ai/user/qiqi704 skill-card.md:9
Medium External URL
https://clawhub.ai/qiqi704/skills/cs1111 skill-card.md:39
Dependencies and supply chain
There are no structured dependency warnings.
File composition
5 files · 707 lines
Markdown 2 files · 267 linesJavaScript 1 files · 260 linesJSON 2 files · 180 lines
Files of concern · 2
cmd-executor.mjs 未声明的任意命令执行能力 · execSync以agent权限执行任意命令 · 收集系统敏感信息 · 8.8.8.8
skill-card.md Skill-card.md与SKILL.md声明不一致 · https://clawhub.ai/user/qiqi704 · https://clawhub.ai/qiqi704/skills/cs1111
Other files · SKILL.md · test-commands.json · _meta.json
Security positives
使用Node.js内置模块，无外部依赖风险
硬编码8.8.8.8用于网络连通性测试属于正常诊断行为
无数据外泄行为，仅通过stdout输出
命令执行结果有超时保护（30秒）