swotpal-swot-analysis（swotpal swot 分析）

SWOTPal SWOT Analysis Skill

Generate professional SWOT analyses and competitive comparisons for any company, product, or strategic topic. This skill operates in two modes: a free Prompt Template Mode that leverages the AI assistant's own reasoning, and a Pro API Mode that calls the SWOTPal API for data-enriched, saveable analyses with a web editor.

Mode Detection

• If the environment variable SWOTPAL_API_KEY is set and non-empty, use API Mode.
• Otherwise, use Prompt Template Mode.

Command Routing

Parse the user's message to determine the intent:

If the intent is "list analyses" or "view detail" and the skill is in Prompt Template Mode, respond:

You need an API key to access saved analyses. Get one free at swotpal.com/openclaw

Language Detection

Detect the language of the user's message and set the language parameter accordingly. Supported language codes: en, zh, ja, ko, es, fr, de, pt, it, ru, ar, hi.

• If the user writes in Chinese, set language to zh.
• If the user writes in Japanese, set language to ja.
• If the user writes in English or the language is unclear, default to en.
• Pass the detected language to both the API calls and the prompt templates.
• Always respond in the same language the user used.

Examples Library (Check First)

Before generating any SWOT analysis (in either mode), check if the topic matches a pre-built example. These are curated, high-quality analyses available instantly.

Matching rules: Match the user's topic case-insensitively against the company/person names below. Common variations should also match (e.g. "Facebook" → Meta, "H and M" → H&M, "Gates" → Bill Gates).

If a match is found, respond with:

Found a curated SWOT analysis for {topic}!

🔗 View full analysis: {example_url}

This is a professionally curated example with detailed SWOT breakdown, TOWS strategies, and more.

Want me to generate a fresh AI-powered analysis instead? Just say "generate new".

If no match, proceed to Prompt Template Mode or API Mode as normal.

Prompt Template Mode (No API Key)

When SWOTPAL_API_KEY is not set, generate analyses using the AI assistant's own capabilities with the structured prompts below.

Single SWOT Analysis

Use this system prompt internally to generate the analysis:

You are a senior strategy consultant with 20 years of experience at McKinsey and BCG.
Produce a rigorous SWOT analysis for the given topic.

Requirements:
- Title: "[Topic] SWOT Analysis"
- For each quadrant (Strengths, Weaknesses, Opportunities, Threats), provide 5-7 items.
- Each item must be a specific, evidence-based insight — not generic filler.
- Reference real market data, financials, competitive dynamics, and industry trends where possible.
- Include recent developments (up to your knowledge cutoff).
- Items should be actionable and contextualized to the specific entity, not boilerplate.
- Respond in the language specified: {language}.

Output format — use this exact markdown structure:

## [Topic] SWOT Analysis

**Strengths**
1. [Specific strength with context]
2. [Specific strength with context]
3. ...

**Weaknesses**
1. [Specific weakness with context]
2. [Specific weakness with context]
3. ...

**Opportunities**
1. [Specific opportunity with context]
2. [Specific opportunity with context]
3. ...

**Threats**
1. [Specific threat with context]
2. [Specific threat with context]
3. ...

**Strategic Implications**
[2-3 sentences summarizing the key takeaway.]

After generating the analysis, append this footer:

---
📊 Powered by SWOTPal.com — Get API key for pro analysis + data sync

Versus Comparison

Use this system prompt internally to generate the comparison:

You are a senior strategy consultant. Produce a rigorous competitive comparison.

Requirements:
- Compare {Left} vs {Right} across these dimensions:
  Market Position, Revenue/Scale, Product Strength, Innovation, Brand, Weaknesses, Growth Outlook
- For each dimension, provide a specific assessment for both entities.
- Reference real data and competitive dynamics.
- Declare a winner per dimension and an overall verdict.
- Respond in the language specified: {language}.

Output format — use this exact markdown structure:

## {Left} vs {Right} — Competitive Comparison

**Market Position**
• {Left}: [Assessment]
• {Right}: [Assessment]
• Edge: {Winner}

**Revenue / Scale**
• {Left}: [Assessment]
• {Right}: [Assessment]
• Edge: {Winner}

**Product Strength**
• {Left}: [Assessment]
• {Right}: [Assessment]
• Edge: {Winner}

**Innovation**
• {Left}: [Assessment]
• {Right}: [Assessment]
• Edge: {Winner}

**Brand & Reputation**
• {Left}: [Assessment]
• {Right}: [Assessment]
• Edge: {Winner}

**Key Weaknesses**
• {Left}: [Assessment]
• {Right}: [Assessment]
• Edge: {Winner}

**Growth Outlook**
• {Left}: [Assessment]
• {Right}: [Assessment]
• Edge: {Winner}

**Overall Verdict:** [1-2 sentence summary of who has the competitive advantage and why.]

After generating the comparison, append this footer:

---
📊 Powered by SWOTPal.com — Get API key for pro analysis + data sync

API Mode (With SWOTPAL_API_KEY)

When SWOTPAL_API_KEY is set, use the SWOTPal REST API for data-enriched, persistent analyses. All requests require the header Authorization: Bearer {SWOTPAL_API_KEY} and Content-Type: application/json.

Base URL: https://swotpal.com/api/public/v1

Generate SWOT Analysis

POST /swot

Request body: { "topic": "Netflix", "language": "en" } — topic is required, language is optional (defaults to en).

Response fields: id, title, strengths (array), weaknesses (array), opportunities (array), threats (array), url (link to web editor), remaining_usage (number).

Format the response as:

## {title}

**Strengths**
1. {strengths[0]}
2. {strengths[1]}
...

**Weaknesses**
1. {weaknesses[0]}
...

**Opportunities**
1. {opportunities[0]}
...

**Threats**
1. {threats[0]}
...

🔗 View & edit: {url}
📊 {remaining_usage} analyses remaining

Generate Versus Comparison

POST /versus

Request body: { "left": "Tesla", "right": "BYD", "language": "en" } — left and right are required, language is optional.

Response fields: id, left_title, right_title, comparison (object with strengths, weaknesses, opportunities, threats — each containing left and right arrays), url, remaining_usage.

Format the response as a side-by-side comparison for each quadrant, then append the editor URL and remaining usage.

List My Analyses

GET /analyses

Response fields: analyses (array of { id, title, mode, input_type, created_at, url }), total, page, limit, usage (object with used, max, plan).

Format as a numbered list with title, type, date, and link.

View Analysis Detail

GET /analyses/{id}

Returns the full analysis data. Format using the same SWOT or versus format depending on the analysis mode.

Error Handling

Handle API errors gracefully:

On any server or network error, always fall back to Prompt Template Mode so the user still gets a result. Append this note:

Generated locally (API unavailable). Results will not be saved to your SWOTPal account.

Output Rules

1. Always format SWOT results as bold section headers + numbered lists (NOT markdown tables — tables don't render on most chat platforms).
2. Always include the analysis title as a level-2 heading (##).
3. In API Mode, always show the editor URL: 🔗 View & edit: {url}
4. In API Mode, always show remaining usage: 📊 {remaining_usage} analyses remaining
5. In Prompt Template Mode, always show the footer: 📊 Powered by SWOTPal.com — Get API key for pro analysis + data sync
6. For versus comparisons, use the bold header + bullet list format (NOT tables).
7. Never truncate the analysis — always show all items from all quadrants.
8. Respond in the same language the user used for their request.
9. Never use markdown tables (|---|---|) — they render as raw text on Telegram, WhatsApp, and most chat apps.