中文 EN
Start Review
Skill Trust Decision

agent-credential-vault

纯文档型skill，声称凭证server-side brokered安全，但无代码验证。凭证实际发送到外部SaaS服务(api.useanima.sh)，文档-行为一致性无法验证。

Install decision first Source: ClawHub Scanned: 1 day ago
Files 1
Artifacts 5
Violations 0
Findings 3

Why this conclusion was reached

0/4 dimensions flagged
Pass
Declared vs actual capability

Declared resources and inferred behavior are broadly aligned.

Review
Hidden execution and egress

5 lower-risk artifacts were extracted and still need context.

Pass
Attack chain and severe findings

There is no explicit malicious chain in the report.

Review
Dependencies and supply chain hygiene

Dependency information is incomplete, so supply-chain confidence stays limited.

What drove the risk score up

凭证外传至第三方SaaS +20

所有凭证通过api.useanima.sh brokered，无代码验证实际行为

纯文档型skill +15

无脚本代码，声明的安全特性无法验证

外部API依赖 +10

完全依赖anima.sh服务，离线或服务被接管则失效

Most important evidence

Medium Doc Mismatch

声明安全性与可验证性差距

SKILL.md声称凭证'server-side brokered'、'never returns plaintext'、'never appears in argv/logs'，但无实际代码验证。这些是核心安全声明却无法验证。

SKILL.md:6
要求提供开源实现或第三方安全审计报告
Medium Data Exfil

凭证发送到外部服务器

vault use命令将凭证发送到https://api.useanima.sh执行HTTP调用。虽然声称server-side，但凭证实际经过第三方服务器。

SKILL.md:5
评估是否信任Anima作为凭证托管方；考虑凭证的敏感性级别
Low Supply Chain

SaaS服务依赖风险

完全依赖anima.sh服务的可用性和安全性。服务被入侵或行为变更时无法察觉。

SKILL.md:4
考虑本地开源替代方案或自托管

Declared capability vs actual capability

Network Pass
Declared READ
→
Inferred WRITE
SKILL.md:5 metadata api_base https://api.useanima.sh
Environment Pass
Declared WRITE
→
Inferred WRITE
SKILL.md:84 vault exec注入凭证到子进程环境变量
Shell Pass
Declared NONE
→
Inferred READ
SKILL.md:76 dry-run显示解析结果，无命令执行

Suspicious artifacts and egress

Medium External URL
https://useanima.sh

SKILL.md:4

Medium External URL
https://docs.useanima.sh

SKILL.md:5

Medium External URL
https://api.useanima.sh

SKILL.md:6

Medium External URL
https://portal.acme.com

SKILL.md:44

Medium External URL
https://api.acme.com/v1/orders

SKILL.md:69

Dependencies and supply chain

There are no structured dependency warnings.

File composition

1 files · 138 lines
Markdown 1 files · 138 lines
Files of concern · 1
SKILL.md Markdown · 138 lines
声明安全性与可验证性差距 · 凭证发送到外部服务器 · SaaS服务依赖风险 · https://useanima.sh · https://docs.useanima.sh · https://api.useanima.sh · https://portal.acme.com · https://api.acme.com/v1/orders

Security positives

文档结构清晰，声称的安全设计思路合理
使用--password-stdin避免凭证出现在argv
提供--allowed-host和--reveal-policy brokered防止凭证滥用
包含审计和redact功能防止意外泄露