Security audit
Security checks for vulnerabilities and agentic risk
This skill runs a local log-summary script that reads a disclosed fixed production-log path and prints aggregate statistics, with no evidence of hidden execution, exfiltration, persistence, or destructive behavior.
Before installing, confirm that `/tmp/agent_v35_production.log` is the log you intend to analyze and sanitize it if it may contain secrets, user data, internal identifiers, or sensitive operational details. Expect a simple console summary rather than continuous monitoring, strategy analytics, visualizations, or exported report files.
描述与代码部分吻合：代码确实分析生产日志、生成统计摘要，并做了 v3.5 与 v3.0 的简单对比，也可用于性能评估的基础数据汇总。但描述明显夸大了能力。代码没有任何持续监控机制、没有实现策略分析逻辑，虽然 stats 中有 strategies 字段，但从未填充或输出；也没有生成测试报告文件或更完整的实验报告产物，只是打印控制台摘要。因此声明的核心能力范围大于实际实现，构成描述与行为不一致。
The README explicitly states that the skill analyzes production deployment logs and defaults to reading /tmp/agent_v35_production.log, but it provides no warning about sensitive data handling, redaction, retention, or access controls. In the context of an agent skill, this can normalize unsafe processing of production telemetry and lead users to expose secrets, identifiers, or operational data in reports.
The skill advertises execution of a report-generation script and documents reading a production log from a fixed path, but it does not declare any tool scope such as file-read permissions. In an agent ecosystem, undeclared file access weakens least-privilege controls and can lead to unintended access to sensitive production data or approval bypasses.
The description field forces a specific language presentation in Chinese, and the rest of the document also assumes Chinese-language use. There is no indication that this is optional, user-selected, or justified as a region-specific skill, which creates a natural-language locale policy concern.
The file's natural-language content, including the module description, comments, error messages, and report output, is written exclusively in Chinese with no indication that language selection is optional or constrained by a justified region-specific requirement. This creates a locale-policy issue because the skill effectively enforces one language for user-facing interaction without opt-in.
SQP-3 applies to all file types and covers language or locale policy violations. The operational description and usage guidance are presented in Chinese only, with no indication of language choice, translation, or user opt-in, which can amount to forcing a specific language on users.
The skill states that it reads from a fixed production-log path in /tmp but does not provide any warning about handling potentially sensitive production data. This omission can cause users or agents to process production logs without understanding confidentiality, retention, or trust implications, especially if the path is assumed safe simply because it is under /tmp.
No suspicious patterns detected.
