Analyzed 60 prompts
- Total tokens across all prompts: 88,622
- Average tokens per prompt: 1,477
- High optimization potential: 7 prompts
- Medium optimization potential: 53 prompts
- Estimated token savings (50% reduction): 44,311
- File: tasks/audio/generate_txt2img_cinematic_directions.md
- Tokens: 2,504
- Redundancy: 34.78%
- Suggested Patterns: Few-shot: Extract examples into few-shot format, Progressive: Use progressive context enhancement
- File: tasks/audio/timestamped_music_video_storyboard.md
- Tokens: 2,238
- Redundancy: 38.31%
- Suggested Patterns: Few-shot: Extract examples into few-shot format, Progressive: Use progressive context enhancement
- File: tasks/audio/storyboard_with_inferred_audio_properties.md
- Tokens: 1,618
- Redundancy: 53.93%
- Suggested Patterns: Few-shot: Extract examples into few-shot format, Progressive: Use progressive context enhancement
- File: tasks/audio/infer_audio_properties_from_lyrics.md
- Tokens: 1,565
- Redundancy: 41.36%
- Suggested Patterns: Few-shot: Extract examples into few-shot format, Progressive: Use progressive context enhancement, Token-budget: Implement token budgeting for analysis
- File: tasks/audio/timestamped_lyrics_audio_descriptives.md
- Tokens: 1,480
- Redundancy: 38.69%
- Suggested Patterns: Few-shot: Extract examples into few-shot format, Progressive: Use progressive context enhancement
- File: tasks/analysis/curate_process_flows_functionalities.md
- Tokens: 1,211
- Redundancy: 39.48%
- Suggested Patterns: Control-flow: Implement explicit control flow pattern, Few-shot: Extract examples into few-shot format, Progressive: Use progressive context enhancement, Token-budget: Implement token budgeting for analysis
- File: tasks/analysis/text_parsing_categorization.md
- Tokens: 943
- Redundancy: 39.32%
- Suggested Patterns: Control-flow: Implement explicit control flow pattern, Few-shot: Extract examples into few-shot format, Progressive: Use progressive context enhancement, Token-budget: Implement token budgeting for analysis
- Progressive: 58 prompts
- Few-shot: 48 prompts
- Token-budget: 32 prompts
- Control-flow: 17 prompts
- Field-based: 10 prompts