Zhongguancun Industry Research

产学研合作调研。当用户要求对某家企业（如百度、阿里、华为、科大讯飞等）做合作调研、产业调研、合作可行性分析、研究方向匹配、对接准备时触发。输出包括：企业画像、研究方向交叉匹配、合作切入点、参会人员建议，以及可选的会议议程。也适用于用户说"帮我看看XX公司能不能合作""调研一下XX"等场景。

Security Scan

The skill's files and runtime instructions are coherent with its stated purpose (company collaboration research and agenda generation) and it does not request unusual installs, credentials, or system access.

Purpose & Capability

The name/description match the SKILL.md and the reference documents: the skill is an instruction-only research/report generator for产学研合作. It requests no binaries, env vars, or config paths — which is proportionate for a web-search-based research assistant.

Instruction Scope

SKILL.md explicitly instructs using a search tool to collect public/company information (business units, research org, key people, technical posture) and to map that to the included 'our-research' directions, then produce a structured report and optional agenda. It does not instruct reading local files, accessing secrets, or calling unexpected external endpoints. Note: it mentions generating Word/PPT 'using appropriate tools' but does not specify which tools or external services — that step could require additional capabilities not declared here and should be verified before enabling those conversions.

Install Mechanism

No install spec and no code files: the skill is instruction-only and therefore low risk from installation (nothing will be written/executed on disk by the skill itself).

Credentials

The skill declares no environment variables, credentials, or config paths. Its information needs are limited to public research and internal reference docs included in the package. The only privacy-related activity is collecting names/roles from public sources — expected for this function.

Persistence & Privilege

always is false and the skill does not request persistent system-wide changes. The skill runs as an ordinary user-invocable skill; default autonomous invocation is unchanged (normal for skills) and not combined with other red flags.

Assessment

This skill appears coherent and low-risk: it uses web search to compile public company profiles and to map them to your institute's research directions. Before installing, consider: (1) confirm whether the agent will have live web/search access (the skill expects that); (2) do not provide any internal/private credentials or confidential documents — the skill does not require them; (3) if you request Word/PPT output, check which conversion/export tool will be used (that step could call an external service or require additional permissions); (4) monitor first outputs for accuracy (public profiles can be stale) and for any prompt asking for unrelated secrets — if it does, stop and revoke the skill. Overall the package is self-consistent and matches its stated purpose.

产学研合作调研

为中关村两院（北京中关村学院 × 中关村人工智能研究院）与外部企业的合作对接提供结构化调研支持。

核心产出

调研报告（必需）→ 会议议程（可选，基于调研结果）

工作流程

第一步：企业画像调研

使用搜索工具，按以下维度收集信息：

基本信息

• 核心业务板块与营收结构
• 研究院/AI实验室组织架构
• 近期战略重点（财报、新闻、发布会）

关键人物

• 分管AI/研究的副总裁级
• 研究院院长/实验室主任
• 与我方研究方向对口的团队负责人
• 格式：姓名 | 职务 | 研究方向/分管领域

技术布局

• 核心AI产品与技术栈
• 开源项目与学术发表方向
• 近期重点投入的技术领域

第二步：研究方向交叉匹配

对照我方三个方向（详见 references/our-research.md），逐一分析匹配度：

我方方向	对方对应领域	匹配度	合作切入点
AI核心	基础模型/优化/算力	高/中/低	具体说明
AI4S	AI+Science/认知智能	高/中/低	具体说明
AI+社会科学	教育/语言/社会计算	高/中/低	具体说明

匹配度为"高"的方向 → 优先推进，安排议程主讲 匹配度为"中"的方向 → 可选推进，视对方兴趣调整 匹配度为"低"的方向 → 暂不安排，备注后续可能性

第三步：合作模式预判

基于匹配结果，给出建议的合作模式（可多选）：

• 联合课题/联合实验室
• 数据/场景共建
• 人才培养/联合培养
• 技术转化/产品化
• 开源共建
• 其他

每种模式附简要可行性与推进建议。

第四步（可选）：生成会议议程

如用户需要，基于调研结果按模板生成议程（详见 references/agenda-template.md）：

• 议程一的主题和主讲人由第二步匹配结果决定
• 对方参会人员由第一步关键人物推导
• 议程三的合作探讨方向由第三步合作模式决定

输出格式

默认输出结构化调研报告（Markdown），包含以上四个部分。如用户要求特定格式（Word、PPT），使用相应工具生成。