Thai Chinese Gov Efficiency

Academic research agent for comparative analysis of governance structures in Thai and Chinese public business schools and their impact on educational efficie...

Security Scan

The skill's instructions and requirements are internally consistent with an academic research agent, but it relies on external scripts/files and subagents that are not included — verify those before running.

Purpose & Capability

The name/description (comparative governance research) match the SKILL.md content: theoretical framework building, mixed-methods design, literature review, data collection, DEA/regression analysis and coordination of subagents. Nothing requested or described is unrelated to that purpose.

Instruction Scope

The instructions tell the agent to run web_search/web_fetch, execute local scripts (scripts/*.py, Rscript), read local reference files (references/*, assets/*) and spawn other subagents. Those actions are coherent for a research workflow, but the skill does not include the referenced scripts/files or define the behavior of the spawn_subagent commands — the agent may attempt to run commands that depend on missing or untrusted artifacts.

Install Mechanism

No install spec or code files are provided (instruction-only). This minimizes installation risk because nothing is downloaded or written by the skill itself.

Credentials

The skill declares no environment variables, credentials, or config paths. The SKILL.md also does not request secrets. This is proportionate to the stated academic research purpose.

Persistence & Privilege

always is false and the skill does not request persistent/privileged system presence. It does instruct spawning subagents and running update commands (clawhub update), which are normal for agent workflows but worth reviewing before enabling autonomous execution.

Assessment

This skill appears to do what it says, but it expects external scripts, local reference files, and other sub-skills that are not included. Before installing or enabling autonomous use: 1) confirm the existence and trustworthiness of scripts referenced in scripts/*.py and R scripts, and of files under references/ and assets/; 2) verify the target subagents (literature-review, data-analysis) are legitimate and permitted to receive your data; 3) ensure the platform commands used (web_fetch, web_search, spawn_subagent, clawhub) are safe in your environment and won't leak sensitive information; 4) run the skill in a sandbox or with restricted agent permissions first; and 5) if you cannot validate the external artifacts, avoid granting broad autonomous execution rights.

中泰高校治理效率研究代理

核心能力

1. 理论框架构建

• 制度理论整合：结合新制度主义理论与资源依赖理论构建分析框架
• 关键维度：

  | 治理维度       | 中国指标                | 泰国指标               |
  |----------------|-------------------------|------------------------|
  | 行政集权度     | 教育部直接管理比例      | 大学自治委员会决策权重 |
  | 财政自主性     | 预算审批层级            | 校级财政自由裁量权     |
  | 人事控制度     | 编制审批制度            | 校长聘用自主权         |
  

• 参考文件：references/theoretical_framework.md（含制度变迁路径图）

2. 研究方法设计

• 混合研究方法：

  # 数据收集协议
  web_search --query "中国双一流高校治理白皮书 site:edu.cn" --count 5
  web_fetch https://public.moe.gov.cn/jytb_ghfz/ghfw/202312/t20231201_1093012.html
  exec python3 scripts/data_extractor.py --country=TH --source=mua
  

• 质量控制：
  • 三角验证：政策文本+院校年报+专家访谈
  • 信效度检验：Cronbach's α >0.7
• 参考文件：references/research_methods.md（含抽样方案模板）

3. 研究流程管理

graph LR
A[文献系统综述] --> B[指标体系构建]
B --> C[数据采集验证]
C --> D[DEA效率测算]
D --> E[回归分析]
E --> F[政策建议生成]

• 里程碑管理：

  {
    "Q3 2024": "完成5所中泰高校深度案例",
    "Q1 2025": "建立治理-效率面板数据库",
    "Q3 2025": "形成SSCI投稿初稿"
  }
  

4. 代理协调机制

• 子代理调度：

  spawn_subagent --skill literature-review --params "{\"keywords\": \"高校治理 中国 泰国\", \"years\": 2019-2024}""
  spawn_subagent --skill data-analysis --tool exec --command "Rscript scripts/efficiency_model.R"
  

• 冲突解决：当数据矛盾时自动触发：

  exec python3 scripts/discrepancy_resolver.py --threshold=0.2
  

质量保障协议

1. 每周执行 clawhub update --skill academic-research 同步最新方法论
2. 所有数据源需通过 references/quality_checklist.md 验证
3. 关键结论需经2名领域专家确认（记录在 assets/peer_review_log/）