Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →核心竞争力提炼助手。专注于分析和提炼公司的核心竞争力/护城河，包括品牌、成本、网络效应、切换成本、牌照壁垒等维度。 **触发场景**： - 用户需要分析公司核心竞争力（"XX 公司的护城河是什么"、"竞争优势在哪里"） - 需要提炼公司投资亮点、竞争壁垒 - 需要竞品对比分析（与对手的优劣势） - 需要写公司研究报告的核心竞争力章节 - 需要客户沟通材料中的公司优势总结 **关键词**："核心竞争力"、"护城河"、"竞争优势"、"壁垒"、"亮点"、"优势"、"竞品对比"、"优劣势"
.claude/skills/aifinlab-competitive-advantage-analysis/SKILL.md| Test case | Without → With | Effect | Δ tokens | Δ turns |
|---|---|---|---|---|
| case-15 | ✗→✓ | ▲ Improved | 109% | 0% |
| case-18 | ✗→✓ | ▲ Improved | 38% | 0% |
| case-23 | ✗→✓ | ▲ Improved | 78% | 0% |
| case-03 | ✓→✗ | ▼ Worse | 26% | 0% |
| case-05 | ✓→✗ | ▼ Worse | 65% | 0% |
你是一名经验丰富的分析师，擅长深入分析并清晰提炼公司的核心竞争力与护城河，帮助投资者理解公司的长期投资价值。
| 护城河类型 | 定义 | 典型公司 | |------------|------|----------| | 品牌护城河 | 品牌认知度高、有溢价能力 | 贵州茅台、苹果 | | 成本护城河 | 规模效应、成本优势 | 宁德时代、海螺水泥 | | 网络护城河 | 用户网络效应、平台效应 | 腾讯、美团 | | 切换护城河 | 客户粘性高、转换成本高 | 微软、用友网络 | | 牌照护城河 | 准入壁垒、政策保护 | 中国中免、券商 |
1. 护城河类型（品牌/成本/网络/切换/牌照）
2. 护城河强度（强/中/弱）
3. 核心优势（1-2 条）
4. 可持续性判断1. 护城河类型识别
2. 护城河强度分析（量化指标）
3. 竞品对比（与对手的优劣势）
4. 护城河变化趋势（加深/变浅）
5. 可持续性判断
6. 投资风险在标准分析基础上增加：
| 维度 | 公司 | 对手 A | 对手 B | 优势/劣势 | |------|------|--------|--------|-----------| | 品牌力 | | | | | | 成本优势 | | | | | | 技术壁垒 | | | | | | 客户粘性 | | | | | | 渠道能力 | | | | | | 规模效应 | | | | |
输出前自查：
# 【宁德时代核心竞争力分析】
## 护城河类型
- 主要护城河：成本护城河 + 技术护城河
- 次要护城河：规模效应 + 客户粘性
## 护城河强度分析
### 成本优势
- 产能规模：2023 年产能 400GWh，全球第一
- 单位成本：比二线厂商低 15%-20%
- 上游布局：锂矿、正极材料自供比例提升
### 技术壁垒
- 研发投入：2023 年研发费用 150 亿，研发人员 2 万+
- 专利数量：累计专利 1.4 万 +，行业第一
- 技术领先：麒麟电池、钠离子电池领先行业 1-2 年
### 客户粘性
- 客户结构：特斯拉、宝马、蔚来等头部车企
- 绑定深度：战略合作 + 合资建厂
- 切换成本：车企认证周期 2-3 年，切换成本高
## 竞品对比
| 维度 | 宁德时代 | 比亚迪 | 中创新航 |
|------|----------|--------|----------|
| 产能 (GWh) | 400 | 300 | 50 |
| 市占率 | 37% | 16% | 5% |
| 研发费用 (亿) | 150 | 120 | 10 |
| 单位成本 | 基准 | +10% | +20% |
## 护城河变化趋势
- 趋势：护城河持续加深
- 驱动：产能扩张 + 技术迭代 + 客户绑定
## 可持续性判断
- 短期（1-2 年）：护城河稳固，领先优势明显
- 中期（3-5 年）：面临比亚迪等对手追赶，但领先优势仍存
- 长期（5 年+）：需关注技术路线变革风险（固态电池等）
## 投资风险
- 技术路线变革（固态电池颠覆）
- 产能过剩导致价格战
- 客户自研电池（特斯拉、蔚来等）| Case | Status | Duration (ms) | Turns | Tokens | Tool calls | ||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Without | With | Δ | Without | With | Δ | Without | With | Δ | Without | With | Δ | ||
case-01 | fail→fail | 12,641 | 17,247 | +36% | 1 | 1 | 0% | 1,914 | 3,559 | +86% | 0 | 0 | — |
case-02 | fail→fail | 44,138 | 51,762 | +17% | 1 | 1 | 0% | 6,502 | 9,579 | +47% | 0 | 0 | — |
case-03 | pass→fail | 24,314 | 23,267 | -4% | 1 | 1 | 0% | 5,113 | 6,436 | +26% | 0 | 0 | — |
case-04 | pass→pass | 12,863 | 13,222 | +3% | 1 | 1 | 0% | 2,279 | 3,889 | +71% | 0 | 0 | — |
case-05 | pass→fail | 21,858 | 24,721 | +13% | 1 | 1 | 0% | 3,242 | 5,353 | +65% | 0 | 0 | — |
case-06 | pass→pass | 22,743 | 22,290 | -2% | 1 | 1 | 0% | 3,481 | 4,916 | +41% | 0 | 0 | — |
case-07 | pass→pass | 20,515 | 24,722 | +21% | 1 | 1 | 0% | 3,270 | 5,575 | +70% | 0 | 0 | — |
case-08 | pass→pass | 22,753 | 21,226 | -7% | 1 | 1 | 0% | 3,282 | 4,656 | +42% | 0 | 0 | — |
case-09 | fail→fail | 20,440 | 21,224 | +4% | 1 | 1 | 0% | 3,044 | 4,509 | +48% | 0 | 0 | — |
case-10 | pass→pass | 21,601 | 22,426 | +4% | 1 | 1 | 0% | 3,250 | 5,021 | +54% | 0 | 0 | — |
case-11 | pass→pass | 21,162 | 23,023 | +9% | 1 | 1 | 0% | 3,162 | 5,261 | +66% | 0 | 0 | — |
case-12 | pass→pass | 22,892 | 26,122 | +14% | 1 | 1 | 0% | 3,249 | 5,176 | +59% | 0 | 0 | — |
case-13 | fail→fail | 19,352 | 18,354 | -5% | 1 | 1 | 0% | 2,676 | 4,266 | +59% | 0 | 0 | — |
case-14 | pass→pass | 44,205 | 40,096 | -9% | 1 | 1 | 0% | 6,018 | 7,421 | +23% | 0 | 0 | — |
case-15 | fail→pass | 11,110 | 11,676 | +5% | 1 | 1 | 0% | 1,609 | 3,360 | +109% | 0 | 0 | — |
case-16 | pass→pass | 25,702 | 22,651 | -12% | 1 | 1 | 0% | 3,845 | 4,855 | +26% | 0 | 0 | — |
case-17 | pass→pass | 23,842 | 32,363 | +36% | 1 | 1 | 0% | 3,550 | 6,292 | +77% | 0 | 0 | — |
case-18 | fail→pass | 27,088 | 25,657 | -5% | 1 | 1 | 0% | 3,944 | 5,447 | +38% | 0 | 0 | — |
case-19 | pass→pass | 18,471 | 21,863 | +18% | 1 | 1 | 0% | 2,668 | 4,908 | +84% | 0 | 0 | — |
case-20 | pass→pass | 23,474 | 23,197 | -1% | 1 | 1 | 0% | 3,363 | 5,056 | +50% | 0 | 0 | — |
case-21 | pass→pass | 24,874 | 25,047 | +1% | 1 | 1 | 0% | 3,621 | 5,271 | +46% | 0 | 0 | — |
case-22 | fail→fail | 23,096 | 26,308 | +14% | 1 | 1 | 0% | 3,310 | 5,471 | +65% | 0 | 0 | — |
case-23 | fail→pass | 25,075 | 32,524 | +30% | 1 | 1 | 0% | 3,627 | 6,469 | +78% | 0 | 0 | — |
DecimalAI ran this skill against gemini-3.6-flash twice over the same eval suite — once with the skill loaded and once without — and compared the two runs case by case. 23 cases were attempted. The headline lift of +4 percentage points is the difference between those two pass rates over the 23 comparable cases. 2 cases got worse with the skill loaded, and they are included in that figure.
Without the skill loaded, the model failed this case. With it loaded, the same prompt on the same model passed. This is one improved case from the latest verified run; every case, including any that regressed, is in the table above.
Other measured skills in the registry, with their headline benchmark lift.