Skill Trust Decision

journal-monitor

monitor.py 是纯良性的 PubMed 文献监控脚本,无恶意代码;但 SKILL.md 包含反安全/提示词注入特征的强约束语言。

Install decision first Source: ClawHub Scanned: Apr 9, 2026
Files 2
Artifacts 1
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

1 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

1 dependency or supply-chain issues need attention.

What drove the risk score up

SKILL.md 反安全/反提示注入语言 +15

使用 "FORCED"、"MANDATORY"、"NO tool-switching" 等强约束词汇,试图限制 LLM 使用更安全的替代工具,符合 prompt_injection 特征但代码本身无恶意

硬编码邮箱地址 +5

Entrez.email = "[email protected]" 在代码中硬编码,标准 NCBI API 要求但可考虑环境变量方式

Most important evidence

Medium Prompt Injection

SKILL.md 包含反安全/反提示注入的强约束语言

SKILL.md 使用 "FORCED translation, MANDATORY metadata, and NO tool-switching" 以及 "MUST ONLY use data returned by monitor.py"、"DO NOT switch to internal search" 等强约束语句,试图阻止 LLM 使用更安全或合法的替代工具,属于 prompt injection 特征但未达到恶意级别。

SKILL.md:1
将强制约束语言改为描述性说明,如"推荐使用 monitor.py 获取数据"而非禁止替代工具
Low Doc Mismatch

邮箱地址硬编码(潜在个人信息暴露)

Entrez.email 硬编码为 [email protected],属于个人邮箱,可能随代码分发而暴露个人信息

monitor.py:8
改用环境变量或 _meta.json 配置,如 os.environ.get('NCBI_EMAIL')
Low Supply Chain

biopython 依赖无版本锁定

metadata 声明 "uv pip install biopython" 无版本约束,可能引入依赖供应链风险

SKILL.md:11
改为版本锁定安装:uv pip install biopython==1.84 或使用 requirements.txt

Declared capability vs actual capability

Network Pass
Declared READ
→
Inferred READ
monitor.py:36 使用 Entrez.esearch/efetch 访问 PubMed API(声明与行为一致)
Filesystem Pass
Declared NONE
→
Inferred NONE
monitor.py 无任何文件读写操作
Shell Pass
Declared NONE
→
Inferred NONE
monitor.py 无 subprocess/os.system 等 shell 调用
Environment Pass
Declared NONE
→
Inferred NONE
monitor.py 仅依赖 CLI 参数和 os.environ(sys.argv),无凭证遍历
Skill Invoke Pass
Declared NONE
→
Inferred NONE
无嵌套 skill 调用

Dependencies and supply chain

PackageVersionSourceKnown vulnNotes
biopython * pip No SKILL.md 中声明为 "uv pip install biopython",无版本锁定,存在供应链风险

File composition

2 files · 124 lines
Python 1 files · 75 linesMarkdown 1 files · 49 lines
Files of concern · 2
monitor.py Python · 75 lines
邮箱地址硬编码(潜在个人信息暴露) · [email protected]
SKILL.md Markdown · 49 lines
SKILL.md 包含反安全/反提示注入的强约束语言 · biopython 依赖无版本锁定

Security positives

monitor.py 代码逻辑完全良性和透明,仅使用 Biopython Entrez API 查询 PubMed
无文件写入、无网络外传、无凭证收割、无 shell 执行
无混淆、Base64 编码或反分析技术
代码结构清晰,易于人工审计
能力声明(network:READ)与实际行为一致