name: prompt-elo description: Get detailed Elo scoring breakdown with visual report for your last prompt allowed-tools:

• Bash(python3 *)
• Bash(open *)
• Write(*)

PromptElo Detailed Analysis

When this skill is invoked, analyze the user's most recent prompt and generate a detailed visual report.

Instructions

1. First, get the user's most recent prompt from the conversation history. If the skill was invoked with arguments (e.g., /prompt-elo "my prompt here"), use that text instead.

2. Run the scorer to get the detailed breakdown:

   echo '{"prompt": "<USER_PROMPT>"}' | python3 ~/.promptelo/client/scorer.py --detailed
   

3. Generate an HTML report using the template at ~/.promptelo/skills/prompt-elo/templates/report.html

4. The report should include:

   • Overall Elo Score with tier badge
   • Radar Chart showing all 5 criteria (clarity, specificity, context, creativity, novelty)
   • Score Breakdown with explanations for each criterion
   • Improvement Suggestions based on lowest-scoring areas
   • Global Ranking showing percentile among all prompts (if API available)

5. Save the report to a temp file and open it in the browser:

   open /tmp/promptelo_report.html
   

Output Format

Present a summary in the terminal:

📊 PromptElo Analysis Complete!

Elo Rating: 1847 ⭐ (Expert)

Score Breakdown:
  Clarity:     ████████░░  82%
  Specificity: ███████░░░  71%
  Context:     █████████░  88%
  Creativity:  ██████░░░░  63%
  Novelty:     ███████░░░  74%

🎯 Top Suggestion: Add more specific technical details to improve specificity.

📈 Global Ranking: Top 15% (of 12,847 prompts)

Full report opened in browser →

Notes

• If the community API is unavailable, novelty will show as "N/A" and use a default value for Elo calculation
• The HTML report includes interactive elements and can be saved for reference
• Suggestions are generated based on which criteria scored lowest