waffles

README

Waffles - LLM Toolchain Orchestrator

Waffles is a command-line tool that orchestrates an LLM toolchain for local development workflows. It automates the process of gathering project context, retrieving prompts, and executing LLM queries, making AI-assisted development more efficient and consistent. Waffles should be pronounced with a Donkey from Shrek accent.

Features

• Automated Pipeline Orchestration: Seamlessly chains wheresmyprompt → files2prompt → llm for complete workflow automation
• Intelligent Repository Analysis: Detects project language, analyzes file structure, and applies smart filtering
• Dependency Management: Automatic installation and verification of required tools
• Interactive Setup: Guided configuration wizard for first-time setup
• Execution History: Persistent logging and querying of all pipeline executions
• Export Capabilities: Multiple output formats (JSON, CSV, Markdown, SQL) for data analysis
• Flexible Configuration: Environment-based configuration with per-project overrides

Quick Start

Installation

# Using Homebrew (recommended)
brew install toozej/tap/waffles

# Using Go
go install github.com/toozej/waffles/cmd/waffles@latest

# From source
git clone https://github.com/toozej/waffles.git
cd waffles
make install

First Run Setup

# Interactive setup wizard
waffles setup

# Or setup with auto-installation
waffles setup --auto-install

Basic Usage

# Execute a simple query
waffles query "explain this code structure"

# Query with specific model
waffles query --model gpt-4 "review this code for security issues"

# Export execution history
waffles export --format json --output results.json

Table of Contents

1. Installation Guide
2. Usage Guide
3. Configuration
4. Commands Reference
5. Troubleshooting
6. Examples

Core Workflow

Waffles orchestrates a three-step pipeline:

1. wheresmyprompt: Retrieves system prompts and context
2. files2prompt: Extracts relevant project files and context
3. llm: Executes the LLM query with combined context

graph LR
    A[Query Input] --> B[wheresmyprompt]
    B --> C[files2prompt]
    C --> D[LLM]
    D --> E[Output]
    
    F[Config] --> B
    F --> C
    F --> D
    
    G[Repository Analysis] --> C

Key Benefits

• Consistency: Standardized workflow ensures consistent results across projects
• Efficiency: Automated context gathering saves time and reduces errors
• History: Track all executions with detailed logging and analytics
• Flexibility: Customizable for different project types and requirements
• Integration: Works with existing development workflows and tools

Requirements

Required Dependencies

• wheresmyprompt: System prompt retrieval tool
• files2prompt: File context extraction tool
• llm: LLM CLI interface

Optional Dependencies

• 1Password CLI: For secure credential management (if using 1Password integration)

All dependencies can be automatically installed during setup.

Quick Examples

Basic Query

waffles query "What does this codebase do?"

Custom File Patterns

waffles query --include "*.go,*.md" --exclude "vendor/*" "Analyze the Go code"

Different Models

waffles query --model claude-3-sonnet "Review for performance issues"
waffles query --model gpt-4 "Suggest architectural improvements"

Export Results

# Export as JSON
waffles export --format json --filter "language=go" --output go-queries.json

# Export as CSV for analysis
waffles export --format csv --days 7 --output weekly-usage.csv

Documentation

• Installation Guide: Detailed installation instructions for all platforms
• Usage Guide: Comprehensive usage examples and best practices
• Configuration: Configuration options and environment variables
• Commands Reference: Complete command reference with all options
• Troubleshooting: Common issues and solutions
• Examples: Real-world usage examples

Support

• Issues: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: docs/

License

See LICENSE for details.

Contributing

Contributions are welcome! Please see CONTRIBUTING.md for guidelines.