メニュー
This skill should be used when the theme-analyzer agent needs guidance on analyzing intelligence cards within a theme, identifying trends, and generating analysis materials.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
This skill should be used when generating intelligence card markdown, creating daily/weekly/monthly reports, formatting agent JSON output, or the user asks to "create an intelligence card", "generate report template", "format report output".
This skill should be used when generating theme reports and panorama reports from analysis materials. Provides markdown templates and structure guidance.
This skill should be used when the user wants to interact with Obsidian from Claude Code. Triggers when the user asks to "search obsidian", "query bases", "read note properties", "set note properties", "create notes in obsidian", or work with Obsidian-specific files like .base files.
This skill should be used when extracting strategic intelligence from documents, determining what information has strategic value, and applying quality standards for intelligence extraction. Triggers on "extract intelligence", "analyze document", "strategic value", or when evaluating information significance.
This skill should be used when analyzing cybersecurity-related documents for strategic intelligence extraction. Triggers when processing documents about threat intelligence, vendor analysis, emerging security technologies, market trends, policy regulations, or capital investment in the security industry.
This skill should be used when the intelligence-cluster agent needs guidance on clustering intelligence cards to themes, detecting new themes, and determining card-theme assignments.
| name | thematic-methodology |
| description | This skill should be used when the theme-analyzer agent needs guidance on analyzing intelligence cards within a theme, identifying trends, and generating analysis materials. |
主题分析方法论，指导如何对已聚类的情报卡片进行深度分析，识别趋势、模式和洞察。
阅读该主题下所有情报卡片，提取：
intelligence_date 分布primary_domain, secondary_domains## 核心事实 章节分析情报的时间分布：
1. 统计每月/每周情报数量
2. 识别增长/下降趋势
3. 检测异常峰值
4. 分析趋势变化原因
分析关键词和实体的热度变化：
1. 统计关键词出现频次
2. 对比早期与近期的关键词变化
3. 识别新兴关键词和消退关键词
构建实体关系网络：
厂商 ↔ 产品 ↔ 技术
威胁组织 ↔ 攻击手法 ↔ 目标行业
投资方 ↔ 被投公司 ↔ 业务领域
分析本主题涉及的领域间关系：
例如 AI 安全主题：
- Emerging-Tech ↔ Threat-Landscape（新技术带来新威胁）
- Vendor-Intelligence ↔ Capital-Investment（厂商融资驱动技术发展）
根据分析维度提炼洞察：
| 维度 | 分析内容 |
|---|---|
| 威胁行为者 | 活跃组织、攻击偏好、演变趋势 |
| 攻击手法 | 新技术、工具变化、防御难点 |
| 目标分析 | 行业偏好、地区分布、规模特征 |
| 影响评估 | 经济损失、数据泄露、业务中断 |
| 维度 | 分析内容 |
|---|---|
| 技术成熟度 | 概念→早期→成长→成熟演变 |
| 应用场景 | 适用行业、部署模式、落地进展 |
| 代表厂商 | 产品能力、市场份额、差异化 |
| 安全影响 | 新攻击面、防御能力提升 |
| 维度 | 分析内容 |
|---|---|
| 市场规模 | 当前规模、增长率、预测数据 |
| 竞争格局 | 主要玩家、市场份额、差异化策略 |
| 客户需求 | 痛点变化、预算趋势、决策因素 |
| 资本动态 | 融资事件、估值变化、投资热点 |
跟踪内容：
跟踪内容：
跟踪内容：
跟踪内容：
跟踪内容：
分析材料输出 JSON：
{
"theme_id": "ai-security",
"theme_name": "AI 安全",
"analysis_date": "2026-03-12T10:30:00Z",
"card_count": 25,
"date_range": {
"start": "2025-06-01",
"end": "2026-03-10"
},
"summary": "主题分析摘要，1-3 句话",
"trends": {
"time_trend": {
"direction": "increasing",
"description": "情报数量呈上升趋势，近3个月增长40%",
"monthly_counts": {"2025-12": 5, "2026-01": 8, "2026-02": 10, "2026-03": 7}
},
"emerging_keywords": ["提示注入", "AI Agent 安全"],
"declining_keywords": ["对抗样本"]
},
"entities": {
"vendors": [
{"name": "OpenAI", "mentions": 8, "key_points": ["发布安全指南", "Red Team 测试"]},
{"name": "Anthropic", "mentions": 5, "key_points": ["Constitutional AI"]}
],
"threat_actors": [],
"technologies": [
{"name": "LLM", "mentions": 15},
{"name": "Prompt Injection", "mentions": 12}
]
},
"cross_domain_links": [
{"domains": ["Emerging-Tech", "Threat-Landscape"], "insight": "AI 技术发展带来新的攻击面"}
],
"key_findings": [
"发现1：AI 安全市场快速增长，年增长率超过 30%",
"发现2：提示注入成为最受关注的 AI 安全威胁",
"发现3：大模型厂商普遍加强安全能力建设"
],
"strategic_implications": [
"建议1：关注 AI 安全产品投资机会",
"建议2：建立 AI 安全测试能力",
"建议3：跟踪监管政策进展"
],
"data_points": [
{"type": "market_size", "value": "50亿美元", "source": "Gartner 报告"},
{"type": "growth_rate", "value": "35% YoY", "source": "行业分析"}
],
"source_cards": ["path/to/card1.md", "path/to/card2.md"]
}
分析完成后执行：
references/dimensions.md - 分析维度详细定义和操作指导