tools

AI Playbook Tools

Utilities for searching, indexing, and optimizing the playbook content.

Quick Start

From the repository root:

# Search for prompts
python tools/search-prompts.py "code review"
python tools/search-prompts.py -t optimization debugging
python tools/search-prompts.py -r "I need to analyze data"

# Rebuild search index after changes
python tools/index-prompts.py

Tools Overview

🚀 adhd-optimizer/

Transforms any prompt into a context-optimized version using ADHD principles:

• 40-60% token reduction while maintaining information fidelity
• Multiple interfaces: Web UI, Python CLI, batch script
• Auto-detects style: technical, debug, learning, creative
• Provides metrics: token counts, clarity scores, structure analysis
• Identifies implicit needs in prompts

Quick usage:

# Command line
python tools/adhd-optimizer/optimize.py "Your long prompt here"

# Interactive mode
python tools/adhd-optimizer/optimize.py -i

# Web interface
open tools/adhd-optimizer/optimizer.html

🔍 search-prompts.py

The main search engine with multiple search modes:

• Keyword search: ./search "your query"
• Tag search: ./search -t tag1 tag2
• Category filter: ./search -c "tasks/coding"
• Archetype search: ./search -a "Clarity Architect"
• Similarity: ./search -s "article writer"
• Recommendations: ./search -r "analyze customer feedback"

🗂️ index-prompts.py

Creates and updates the searchable index of all prompts:

• Extracts metadata from YAML frontmatter
• Builds searchable descriptions
• Tracks tags, categories, and archetypes
• Outputs to prompt-index.json

📊 context-analyzer.py

Analyzes prompts for context engineering optimization opportunities:

• Calculates token usage and redundancy
• Identifies optimization patterns
• Suggests context engineering techniques
• Generates optimization reports

Usage:

# Analyze all prompts
python3 tools/context-analyzer.py

# Analyze specific category
python3 tools/context-analyzer.py -c tasks/writing

# Export detailed JSON report
python3 tools/context-analyzer.py -j analysis.json

# Save report to file
python3 tools/context-analyzer.py -o optimization-report.md

📊 Generated Files

• prompt-index.json: Searchable metadata for all prompts
• prompt-dashboard.md: Quick reference guide

Search Examples

# Find code-related prompts
./search "code"
./search -t coding optimization

# Get writing help
./search "article"
./search -c "tasks/writing"

# Explore vibecoding
./search --list-archetypes
./search -a "Truth Builder"

# Find similar prompts
./search -s "data analysis"

# Get task recommendations
./search -r "I need to debug Python code"

# Detailed results with descriptions
./search -v "machine learning"

Advanced Usage

Combining Searches

Use the search results to build prompt combinations:

1. Find base prompt: ./search "code review"
2. Add framework: ./search "METRICS"
3. Enhance with archetype: ./search -a "Truth Builder"

Maintenance

• Run ./update-index after adding/modifying prompts
• Check prompt-index.json for indexing issues
• Use -v flag to verify descriptions are being extracted correctly

Search Modes Explained

Mode	Flag	Purpose	Example
Keyword	(default)	Full-text search	./search "optimization"
Tags	-t	Match specific tags	./search -t debugging performance
Category	-c	Filter by directory	./search -c "frameworks"
Archetype	-a	Vibecoding search	./search -a "Clarity"
Similar	-s	Find related prompts	./search -s "article writer"
Recommend	-r	Task-based suggestions	./search -r "analyze logs"

Integration with Git

Consider adding these to your git hooks:

# .git/hooks/post-merge
#!/bin/bash
./update-index

# .git/hooks/post-checkout
#!/bin/bash
./update-index

This ensures the search index stays current with repository changes.