Skill Trust Decision

taskfuel

SKILL.md 未声明远程脚本执行的潜在风险,且安全 mitigation(用户确认安装)在文档中缺失

Install decision first Source: ClawHub Scanned: Aug 11, 2026
Files 2
Artifacts 5
Violations 2
Findings 3
Most direct threat evidence
01
用户部署 skill,读取 SKILL.md 理解功能 Entry · SKILL.md
02
skill 自动执行 curl|sh 远程脚本安装 CLI Escalation · SKILL.md
03
恶意安装脚本可在系统植入后门、窃取凭证或建立持久化 Impact · https://taskfuel.ai/install.sh

Why this conclusion was reached

3/4 dimensions flagged
Block
Declared vs actual capability

2 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.

Pass
Dependencies and supply chain hygiene

Dependencies are present but no obvious high-risk issue stands out.

Attack Chain

01
用户部署 skill,读取 SKILL.md 理解功能

Entry · SKILL.md:1

02
skill 自动执行 curl|sh 远程脚本安装 CLI

Escalation · SKILL.md:22

03
恶意安装脚本可在系统植入后门、窃取凭证或建立持久化

Impact · https://taskfuel.ai/install.sh

What drove the risk score up

未声明的远程脚本执行 +25

SKILL.md:22 包含 curl|sh 远程执行但文档未声明此风险,也未说明安装脚本会修改系统

文档与实际行为差异 +15

skill-card.md 提及安装需用户确认,但 SKILL.md 中未体现此安全建议

Most important evidence

High Supply Chain

远程脚本无条件执行

SKILL.md 第 22 行包含 `curl -fsSL https://taskfuel.ai/install.sh | sh`,从远程 URL 下载并直接执行安装脚本,这是经典的供应链攻击向量。脚本内容未经用户审查即执行,可能植入后门或恶意代码。

SKILL.md:22
建议:(1) 将安装命令改为两步:先 curl 下载脚本审查内容,再执行;(2) 在 SKILL.md 顶部明确声明此为高危操作需用户确认;(3) 提供 hash 验证选项
Medium Doc Mismatch

文档缺少安全 mitigation 声明

skill-card.md 明确列出安装需用户确认的 mitigation,但 SKILL.md 作为核心文档未体现此要求,存在文档不一致。

SKILL.md:1
在 SKILL.md 第 0 节前添加风险提示:'⚠️ 警告:安装命令会从远程URL下载并执行脚本,请确保来源可信'
Low Priv Escalation

CLI 安装可能需要 sudo 权限

安装脚本可能向 PATH 添加可执行文件,在某些系统上可能需要提升权限,但文档未说明。

SKILL.md:24
文档应说明可能的权限需求

Declared capability vs actual capability

Filesystem Pass
Declared NONE
Inferred READ
SKILL.md:24 读取配置文件 ~/.config/taskfuel/config.json
Network Block
Declared READ
Inferred WRITE
SKILL.md:22 远程脚本下载+SKILL.md:36+60 API调用
Shell Block
Declared NONE
Inferred WRITE
SKILL.md:22 远程脚本无条件执行

Suspicious artifacts and egress

Critical Dangerous Command
curl -fsSL https://taskfuel.ai/install.sh | sh

SKILL.md:22

Medium External URL
https://taskfuel.ai/install.sh

SKILL.md:22

Medium External URL
https://app.taskfuel.ai

SKILL.md:117

Medium External URL
https://clawhub.ai/user/taskfuel.ai

skill-card.md:9

Medium External URL
https://clawhub.ai/taskfuel.ai/skills/taskfuel

skill-card.md:39

Dependencies and supply chain

PackageVersionSourceKnown vulnNotes
taskfuel CLI unknown remote install script No 安装脚本从远程URL下载,内容未经验证

File composition

2 files · 186 lines
Markdown 2 files · 186 lines
Files of concern · 2
SKILL.md Markdown · 127 lines
远程脚本无条件执行 · 文档缺少安全 mitigation 声明 · CLI 安装可能需要 sudo 权限 · curl -fsSL https://taskfuel.ai/install.sh | sh · https://taskfuel.ai/install.sh · https://app.taskfuel.ai
skill-card.md Markdown · 59 lines
https://clawhub.ai/user/taskfuel.ai · https://clawhub.ai/taskfuel.ai/skills/taskfuel

Security positives

提供 dry-run/quote 机制防止意外扣费
明确消费上限控制(--max-amount)
账户连接需要浏览器用户授权
有明确的错误处理和余额检查