中级Agent Skill

竞品拆解

用有来源的评分卡、价格分析、SWOT 和 UX 审查比较竞品。

浏览量
0
复制次数
0
收藏
0
编辑评分
78/100
支持平台
Claude Code · Codex
01

这个 Skill 是做什么的

用有来源的评分卡、价格分析、SWOT 和 UX 审查比较竞品。文档提供 12 维评分规则和汇报模板,Python 辅助工具可根据 JSON 构建矩阵。

02

它有什么特色

01

方法与优势

文档提供 12 维评分规则和汇报模板,Python 辅助工具可根据 JSON 构建矩阵。

02

适用任务

制作附证据的竞品评分卡;比较定价模式与客户取舍;将评价中的抱怨整理为应对优先级。

03

安装前的特殊要求

需要准备的输入

两到四个竞品、带日期的价格、评价、产品证据和评分假设。

Agent 与运行环境

使用支持 Skill 且能访问文件的 Agent。随包 Python 辅助工具需要 Python 3;提供数据前先查看样例结构。

需确认的适用边界

工具不会采集竞品数据。其可配置数值矩阵与文档的 1–5 分规则不同,比较总分前需统一量表。

04

安装与开始使用

Codex

仓库安装器会把此 Skill 的完整目录复制到 Codex skills 目录。需具备 Git 及安装器要求的 shell/Python 环境。

  1. 按实际 Agent 选择方式,安装完整包,不要只复制页面中的 SKILL.md。
  2. 两到四个竞品、带日期的价格、评价、产品证据和评分假设。
  3. 开启新 Agent 会话,明确要求使用 competitive-teardown。先用小型样本尝试下方一个场景,并检查输出。
git clone https://github.com/alirezarezvani/claude-skills.git cd claude-skills git checkout 6d2f02ea077eead1bee5411a31bcc6db5a11a70c ./scripts/codex-install.sh --skill competitive-teardown
05

原始 Skill.md

以下为已核对版本未经改写的 SKILL.md。安装完整包才能保留引用的脚本、素材和指引。下方启动提示词为 EcomAgentTools 编写的示例。

2026 年 10 月 25 日核对源码及随包文件:GitHub 6d2f02ea077e。本次为静态审查,不代表已完成安装运行或业务效果测试。

查看已核对的源码 · MIT

许可声明
MIT License

Copyright (c) 2025 Alireza Rezvani

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
---
name: "competitive-teardown"
description: "Analyzes competitor products and companies by synthesizing data from pricing pages, app store reviews, job postings, SEO signals, and social media into structured competitive intelligence. Produces feature comparison matrices scored across 12 dimensions, SWOT analyses, positioning maps, UX audits, pricing model breakdowns, action item roadmaps, and stakeholder presentation templates. Use when conducting competitor analysis, comparing products against competitors, researching the competitive landscape, building battle cards for sales, preparing for a product strategy or roadmap session, responding to a competitor's new feature or pricing change, or performing a quarterly competitive review."
---

# Competitive Teardown

**Tier:** POWERFUL  
**Category:** Product Team  
**Domain:** Competitive Intelligence, Product Strategy, Market Analysis

---

## When to Use

- Before a product strategy or roadmap session
- When a competitor launches a major feature or pricing change
- Quarterly competitive review
- Before a sales pitch where you need battle card data
- When entering a new market segment

---

## Teardown Workflow

Follow these steps in sequence to produce a complete teardown:

1. **Define competitors** — List 2–4 competitors to analyze. Confirm which is the primary focus.
2. **Collect data** — Use `references/data-collection-guide.md` to gather raw signals from at least 3 sources per competitor (website, reviews, job postings, SEO, social).  
   _Validation checkpoint: Before proceeding, confirm you have pricing data, at least 20 reviews, and job posting counts for each competitor._
3. **Score using rubric** — Apply the 12-dimension rubric below to produce a numeric scorecard for each competitor and your own product.  
   _Validation checkpoint: Every dimension should have a score and at least one supporting evidence note._
4. **Generate outputs** — Populate the templates in `references/analysis-templates.md` (Feature Matrix, Pricing Analysis, SWOT, Positioning Map, UX Audit).
5. **Build action plan** — Translate findings into the Action Items template (quick wins / medium-term / strategic).
6. **Package for stakeholders** — Assemble the Stakeholder Presentation using outputs from steps 3–5.

---

## Data Collection Guide

> Full executable scripts for each source are in `references/data-collection-guide.md`. Summaries of what to capture are below.

### 1. Website Analysis

Key things to capture:
- Pricing tiers and price points
- Feature lists per tier
- Primary CTA and messaging
- Case studies / customer logos (signals ICP)
- Integration logos
- Trust signals (certifications, compliance badges)

### 2. App Store Reviews

Review sentiment categories:
- **Praise** → what users love (defend / strengthen these)
- **Feature requests** → unmet needs (opportunity gaps)
- **Bugs** → quality signals
- **UX complaints** → friction points you can beat them on

**Sample App Store query (iTunes Search API):**
```
GET https://itunes.apple.com/search?term=<competitor_name>&entity=software&limit=1
# Extract trackId, then:
GET https://itunes.apple.com/rss/customerreviews/id=<trackId>/sortBy=mostRecent/json?l=en&limit=50
```
Parse `entry[].content.label` for review text and `entry[].im:rating.label` for star rating.

### 3. Job Postings (Team Size & Tech Stack Signals)

Signals from job postings:
- **Engineering volume** → scaling vs. consolidating
- **Specific tech mentions** → stack (React/Vue, Postgres/Mongo, AWS/GCP)
- **Sales/CS ratio** → product-led vs. sales-led motion
- **Data/ML roles** → upcoming AI features
- **Compliance roles** → regulatory expansion

### 4. SEO Analysis

SEO signals to capture:
- Top 20 organic keywords (intent: informational / navigational / commercial)
- Domain Authority / backlink count
- Blog publishing cadence and topics
- Which pages rank (product pages vs. blog vs. docs)

### 5. Social Media Sentiment

Capture recent mentions via Twitter/X API v2, Reddit, or LinkedIn. Look for recurring praise, complaints, and feature requests. See `references/data-collection-guide.md` for API query examples.

---

## Scoring Rubric (12 Dimensions, 1-5)

| # | Dimension | 1 (Weak) | 3 (Average) | 5 (Best-in-class) |
|---|-----------|----------|-------------|-------------------|
| 1 | **Features** | Core only, many gaps | Solid coverage | Comprehensive + unique |
| 2 | **Pricing** | Confusing / overpriced | Market-rate, clear | Transparent, flexible, fair |
| 3 | **UX** | Confusing, high friction | Functional | Delightful, minimal friction |
| 4 | **Performance** | Slow, unreliable | Acceptable | Fast, high uptime |
| 5 | **Docs** | Sparse, outdated | Decent coverage | Comprehensive, searchable |
| 6 | **Support** | Email only, slow | Chat + email | 24/7, great response |
| 7 | **Integrations** | 0-5 integrations | 6-25 | 26+ or deep ecosystem |
| 8 | **Security** | No mentions | SOC2 claimed | SOC2 Type II, ISO 27001 |
| 9 | **Scalability** | No enterprise tier | Mid-market ready | Enterprise-grade |
| 10 | **Brand** | Generic, unmemorable | Decent positioning | Strong, differentiated |
| 11 | **Community** | None | Forum / Slack | Active, vibrant community |
| 12 | **Innovation** | No recent releases | Quarterly | Frequent, meaningful |

**Example completed row** (Competitor: Acme Corp, Dimension 3 – UX):

| Dimension | Acme Corp Score | Evidence |
|-----------|----------------|---------|
| UX | 2 | App Store reviews cite "confusing navigation" (38 mentions); onboarding requires 7 steps before TTFV; no onboarding wizard; CC required at signup. |

Apply this pattern to all 12 dimensions for each competitor.

---

## Templates

> Full template markdown is in `references/analysis-templates.md`. Abbreviated reference below.

### Feature Comparison Matrix

Rows: core features, pricing tiers, platform capabilities (web, iOS, Android, API).  
Columns: your product + up to 3 competitors.  
Score each cell 1–5. Sum to get total out of 60.  
**Score legend:** 5=Best-in-class, 4=Strong, 3=Average, 2=Below average, 1=Weak/Missing

### Pricing Analysis

Capture per competitor: model type (per-seat / usage-based / flat rate / freemium), entry/mid/enterprise price points, free trial length.  
Summarize: price leader, value leader, premium positioning, your position, and 2–3 pricing opportunity bullets.

### SWOT Analysis

For each competitor: 3–5 bullets per quadrant (Strengths, Weaknesses, Opportunities for us, Threats to us). Anchor every bullet to a data signal (review quote, job posting count, pricing page, etc.).

### Positioning Map

2x2 axes (e.g., Simple ↔ Complex / Low Value ↔ High Value). Place each competitor and your product. Bubble size = market share or funding. See `references/analysis-templates.md` for ASCII and editable versions.

### UX Audit Checklist

Onboarding: TTFV (minutes), steps to activation, CC-required, onboarding wizard quality.  
Key workflows: steps, friction points, comparative score (yours vs. theirs).  
Mobile: iOS/Android ratings, feature parity, top complaint and praise.  
Navigation: global search, keyboard shortcuts, in-app help.

### Action Items

| Horizon | Effort | Examples |
|---------|--------|---------|
| Quick wins (0–4 wks) | Low | Add review badges, publish comparison landing page |
| Medium-term (1–3 mo) | Moderate | Launch free tier, improve onboarding TTFV, add top-requested integration |
| Strategic (3–12 mo) | High | Enter new market, build API v2, achieve SOC2 Type II |

### Stakeholder Presentation (7 slides)

1. **Executive Summary** — Threat level (LOW/MEDIUM/HIGH/CRITICAL), top strength, top opportunity, recommended action
2. **Market Position** — 2x2 positioning map
3. **Feature Scorecard** — 12-dimension radar or table, total scores
4. **Pricing Analysis** — Comparison table + key insight
5. **UX Highlights** — What they do better (3 bullets) vs. where we win (3 bullets)
6. **Voice of Customer** — Top 3 review complaints (quoted or paraphrased)
7. **Our Action Plan** — Quick wins, medium-term, strategic priorities; Appendix with raw data

## Related Skills

- **Product Strategist** (`product-team/product-strategist/`) — Competitive insights feed OKR and strategy planning
- **Landing Page Generator** (`product-team/landing-page-generator/`) — Competitive positioning informs landing page messaging
06

开始使用

以下是主要功能的用例提示词,可以直接复制使用。

01

制作附证据的竞品评分卡

制作附证据的竞品评分卡 适用于:制作附证据的竞品评分卡。请准备:两到四个竞品、带日期的价格、评价、产品证据和评分假设。

查看 Prompt 和变量收起 Prompt 和变量
请使用已安装的 competitive-teardown Skill。任务:制作附证据的竞品评分卡。输入:[INPUT]。限制:[CONSTRAINTS]。先检查缺失信息,再输出可审查的结果、依据、假设和下一步。工具不会采集竞品数据。其可配置数值矩阵与文档的 1–5 分规则不同,比较总分前需统一量表。

需要替换的变量

[INPUT]
两到四个竞品、带日期的价格、评价、产品证据和评分假设。
[CONSTRAINTS]
任务范围、允许的操作、输出格式和审核要求。
02

比较定价模式与客户取舍

比较定价模式与客户取舍 适用于:比较定价模式与客户取舍。请准备:两到四个竞品、带日期的价格、评价、产品证据和评分假设。

查看 Prompt 和变量收起 Prompt 和变量
请使用已安装的 competitive-teardown Skill。任务:比较定价模式与客户取舍。输入:[INPUT]。限制:[CONSTRAINTS]。先检查缺失信息,再输出可审查的结果、依据、假设和下一步。工具不会采集竞品数据。其可配置数值矩阵与文档的 1–5 分规则不同,比较总分前需统一量表。

需要替换的变量

[INPUT]
两到四个竞品、带日期的价格、评价、产品证据和评分假设。
[CONSTRAINTS]
任务范围、允许的操作、输出格式和审核要求。
03

将评价中的抱怨整理为应对优先级

将评价中的抱怨整理为应对优先级 适用于:将评价中的抱怨整理为应对优先级。请准备:两到四个竞品、带日期的价格、评价、产品证据和评分假设。

查看 Prompt 和变量收起 Prompt 和变量
请使用已安装的 competitive-teardown Skill。任务:将评价中的抱怨整理为应对优先级。输入:[INPUT]。限制:[CONSTRAINTS]。先检查缺失信息,再输出可审查的结果、依据、假设和下一步。工具不会采集竞品数据。其可配置数值矩阵与文档的 1–5 分规则不同,比较总分前需统一量表。

需要替换的变量

[INPUT]
两到四个竞品、带日期的价格、评价、产品证据和评分假设。
[CONSTRAINTS]
任务范围、允许的操作、输出格式和审核要求。
07

风险与使用注意事项

已知限制

工具不会采集竞品数据。其可配置数值矩阵与文档的 1–5 分规则不同,比较总分前需统一量表。

证据与复核

2026 年 10 月 25 日核对源码及随包文件:GitHub 6d2f02ea077e。本次为静态审查,不代表已完成安装运行或业务效果测试。

采用结果前

使用获授权输入,不把密钥贴入提示词。采用前检查面向客户的声明、计算、代码改动及任何拟议的外部操作。

常见问题

它能做什么,有什么优势?

用有来源的评分卡、价格分析、SWOT 和 UX 审查比较竞品。文档提供 12 维评分规则和汇报模板,Python 辅助工具可根据 JSON 构建矩阵。

如何安装和开始使用?

按实际 Agent 选择方式,安装完整包,不要只复制页面中的 SKILL.md。 两到四个竞品、带日期的价格、评价、产品证据和评分假设。 开启新 Agent 会话,明确要求使用 competitive-teardown。先用小型样本尝试下方一个场景,并检查输出。

适合哪些使用场景?

制作附证据的竞品评分卡;比较定价模式与客户取舍;将评价中的抱怨整理为应对优先级

使用前需要了解哪些限制?

工具不会采集竞品数据。其可配置数值矩阵与文档的 1–5 分规则不同,比较总分前需统一量表。

本页核对了什么?

2026 年 10 月 25 日核对源码及随包文件:GitHub 6d2f02ea077e。本次为静态审查,不代表已完成安装运行或业务效果测试。

社区评分结果

任务有效性
安装与使用难度
稳定性与边界
文档清晰度
首次获得有效结果

暂无评论。

发表评论

请先为每一项打分,再分享你的真实使用体验。

任务有效性
安装与使用难度
稳定性与边界
文档清晰度
首次获得有效结果
作者 / 维护者

Alireza Rezvani

2026 年 10 月 25 日核对源码及随包文件:GitHub 6d2f02ea077e。本次为静态审查,不代表已完成安装运行或业务效果测试。

内容核验日期: