You are an SEO research assistant.
Given the following seed keywords: 

Generate a list of 50 related search keywords.
Include variations, synonyms, long-tail keywords, and trending phrases.
Output as JSON with fields:
- keyword
- search_intent (informational / navigational / commercial / transactional)
- reason_for_inclusion

Example output format:
[
  {
    "keyword": "best AI tools for business",
    "search_intent": "commercial",
    "reason_for_inclusion": "High commercial intent, targets business users"
  }
]

Cluster the following keywords into topical groups based on semantic similarity.

Input: {{STEP_1_OUTPUT}}

Output format:
[
  {
    "cluster_name": "AI Productivity Tools",
    "keywords": ["best AI tools for business", "AI tools for productivity", "AI tools 2025"],
    "primary_keyword": "AI productivity tools",
    "search_volume_estimate": "high/medium/low"
  }
]

For each keyword cluster, estimate SEO metrics and content recommendations.

Input: {{STEP_2_OUTPUT}}

For each cluster, provide:
- search_volume_score (1-10)
- ranking_difficulty (1-10, where 10 is hardest)
- trend_score (increasing/stable/decreasing)
- best_content_type (blog, comparison, tutorial, video, infographic, etc.)
- target_audience
- recommended_content_angle

Output as enhanced JSON with all original fields plus new metrics.

Using the enriched keyword data, create a prioritized keyword strategy.
Rank clusters by: high search volume, low difficulty, and strong trend.

Input: {{STEP_3_OUTPUT}}

Output as markdown table with columns:
| Priority | Cluster Name | Keywords | Primary Keyword | Volume | Difficulty | Trend | Content Type | Target Audience | Quick Win Strategy |

After the table, provide:
1. Top 5 Quick Wins (high volume, low difficulty)
2. Long-term Targets (high volume, high difficulty - build authority first)
3. Content Calendar Suggestions (what to create first)
4. Internal Linking Strategy