OpenRouter Interface

A comprehensive Python toolkit for working with AI language models through the OpenRouter API. Supports single prompt processing, multi-prompt chaining, file conversion (docx/pdf/epub), web interface, and advanced automation with pre/post processing scripts.

---

Key Features

• Multiple Interfaces: CLI, Web UI, and programmatic access
• Prompt Chaining: Sequential prompt execution with restart/recovery capabilities
• Smart Recovery: Automatic restart from failed steps with status tracking
• 100+ AI Models: Support for all OpenRouter-compatible models
• Script Integration: Pre/post processing scripts at global and per-step levels
• File Conversion: Convert between markdown and docx, pdf, epub, and more
• Combine Outputs: Merge all step outputs into a single file
• Chapter Splitter: Split documents by chapters, prologues, and epilogues (supports markdown, plain text, Roman numerals)
• Advanced Configuration: YAML-based configuration with parameter overrides
• File Management: Automatic chunking, validation, and format handling
• Web Dashboard: Real-time monitoring and visual chain builder
• Direct File Loading: Upload and execute JSON prompts and YAML chains instantly
• No-Registry Execution: Run prompts and chains without permanent storage

---

Table of Contents

• Quick Start
• Installation
  • Global Installation
  • Local Development
  • Platform-Specific Setup
• Usage
  • Single Prompts
  • Web Interface
  • Prompt Chaining
  • Chapter Splitter
• Configuration
• Examples
• Documentation
• Contributing
• Support

---

Quick Start

5-Minute Setup

# 1. Clone and install
git clone https://github.com/your-org/openrouter-interface.git
cd openrouter-interface
./install-global.sh

# 2. Set up API key
export OPENROUTER_API_KEY="your-api-key-here"

# 3. Run a single prompt
openrouter-runner -p prompts/example.json -i input.md -o output.md

# 4. Create and run a chain
openrouter-chain --create-sample
openrouter-chain -c sample_chain.yaml

# 5. Split a document by chapters
split-chapters book.md -o chapters/

# 6. Start web interface
openrouter-web

---

Installation

Global Installation (Recommended)

Install system-wide for use from anywhere:

git clone https://github.com/your-org/openrouter-interface.git
cd openrouter-interface
./install-global.sh

# Verify installation
openrouter-runner --help
openrouter-web --help
openrouter-chain --help
split-chapters --help

Note: The installer automatically detects and installs pandoc for file conversion features. If automatic installation fails, install manually from pandoc.org.

Local Development

For development or isolated environments:

./install.sh web
source openrouter-venv/bin/activate
PYTHONPATH=src python3 -m openrouter_interface.cli --help

Platform-Specific Setup

Linux

# Ubuntu/Debian
sudo apt update
sudo apt install python3 python3-pip git

# Clone and install
git clone https://github.com/your-org/openrouter-interface.git
cd openrouter-interface
./install-global.sh

# Set API key
echo 'export OPENROUTER_API_KEY="your-api-key-here"' >> ~/.bashrc
source ~/.bashrc

macOS

# Using Homebrew
brew install python git

# Clone and install
git clone https://github.com/your-org/openrouter-interface.git
cd openrouter-interface
./install-global.sh

# Set API key
echo 'export OPENROUTER_API_KEY="your-api-key-here"' >> ~/.zshrc
source ~/.zshrc

Windows

# Install Python 3.8+ and Git
# Clone repository
git clone https://github.com/your-org/openrouter-interface.git
cd openrouter-interface

# Run PowerShell as administrator
powershell -ExecutionPolicy Bypass -File install-windows.ps1

# Set API key
setx OPENROUTER_API_KEY "your-api-key-here"

API Key Setup

Get your API key from OpenRouter and set it up:

# Method 1: Environment variable
export OPENROUTER_API_KEY="your-api-key-here"

# Method 2: Setup script
./setup-api-key.sh

# Method 3: Add to shell profile
echo 'export OPENROUTER_API_KEY="your-api-key-here"' >> ~/.bashrc

---

Usage

Single Prompts

Execute individual prompts against AI models:

# Basic usage
openrouter-runner -p prompts/analysis.json -i document.md -o result.md

# With model and parameter overrides
openrouter-runner -p prompts/creative.json -i input.md -o output.md \
  --model "openai/gpt-4-turbo" \
  --temperature 0.8 \
  --max-tokens 25000

# Multiple prompts in sequence
openrouter-runner -p "prompts/grammar.json,prompts/style.json" \
  -i draft.md -o final.md

Web Interface

Launch the web dashboard for visual prompt management:

# Start with default settings
openrouter-web

# Custom port and debug mode
openrouter-web --port 8080 --debug

# With custom configuration
openrouter-web --config config/web_config.yaml

Web Features:

• File Upload: Drag-and-drop interface
• Model Selection: Choose from 100+ models
• Chain Builder: Visual chain creation
• Real-time Monitoring: Live progress tracking
• File Management: Browse intermediate results
• Load Prompt: Upload and execute JSON prompts instantly
• Load Chain: Upload and execute YAML chains directly

Load Prompt Button

Access from the main page with the green "Load Prompt" button:

1. Upload JSON Prompt: Select any JSON prompt file
2. Preview: View prompt details (title, description, instructions, examples)
3. Configure: Optionally upload YAML configuration for API parameters
4. Input: Provide text or upload file
5. Execute: Run immediately and view results in modal

# Example usage flow:
# 1. Click "Load Prompt" on main page
# 2. Upload: test_load_prompt.json
# 3. Configure: test_single_prompt_config.yaml (optional)
# 4. Input: "This is test content to analyze"
# 5. Execute and view results

Load Chain Configuration (Expanded)

Access from the Chain Runner page - navigates to dedicated full-page interface:

Dedicated Load Page (/chains/load)

1. Navigate: Click "Load Chain" to go to comprehensive load interface
2. Upload Options:
   • File upload with drag-and-drop support
   • Text area for direct YAML input
   • Configuration-based upload with template selection
3. Rich Configuration Analysis:
   • Detailed breakdown showing prompts, models, and parameters
   • Raw YAML preview with syntax highlighting
   • Template usage detection and description
4. Advanced Execution Options:
   • Multiple input methods (text input, file upload, config-specified)
   • Custom output filename overrides
   • Debug mode toggle for detailed logging
   • Validation checks before execution
5. Template Management: Access to pre-built configuration templates
6. Direct Execution: Start chains immediately from the load interface

YAML Editor Integration (/chains/yaml-editor)

• Professional code editor with full YAML syntax highlighting
• Five built-in templates: Basic, Advanced, Scripts, Multi-pass, Append
• Real-time validation with error detection and reporting
• Auto-save functionality and keyboard shortcuts (Ctrl+S, Ctrl+Shift+F)
• Light and dark themes for comfortable editing
• Direct execution: Save & Execute for immediate chain deployment

Enhanced Features:

• Configuration Validation: Pre-execution YAML and config validation
• Error Handling: Comprehensive error reporting with specific line numbers
• Template System: Quick-start templates for common workflow patterns
• Seamless Integration: Edit, validate, and execute in unified interface

# Complete workflow example:
# 1. Click "Load Chain" on /chains page → Full-page load interface
# 2. Upload: complex_chain.yaml → Rich analysis shows 5 prompts, 3 models
# 3. Preview: Raw YAML with syntax highlighting, template detection
# 4. Configure: Custom input method, debug mode enabled
# 5. Validate: Check for YAML errors and missing files
# 6. Execute: Start chain with comprehensive monitoring
# 7. Edit: Jump to YAML editor for quick modifications

Prompt Chaining

Chain multiple prompts for complex workflows:

# Create sample configuration
openrouter-chain --create-sample

# Run a chain
openrouter-chain -c chain_config.yaml

# With debug output
openrouter-chain -c chain_config.yaml --debug

Basic Chain Configuration:

global_config:
  model: "anthropic/claude-4-sonnet-20250522"
  temperature: 0.7

input_file: document.md
output_file: processed_document.md

prompts:
  prompt 1:
    name: "grammar_check"
    prompt_file: "prompts/grammar.json"
    temperature: 0.2

  prompt 2:
    name: "style_improvement"
    prompt_file: "prompts/style.json"
    temperature: 0.8

Chain Restart and Recovery:

OpenRouter Interface includes powerful restart functionality to recover from failures and save time/costs by avoiding re-execution of completed steps.

# Check execution status
openrouter-chain -c config.yaml --status-only

# Restart from failed steps automatically
openrouter-chain -c config.yaml --restart

# Force restart from specific step
openrouter-chain -c config.yaml --restart-from 4

# Clean status and start fresh
openrouter-chain -c config.yaml --clean-status

Key Benefits:

• Time Savings: Skip expensive LLM calls for completed steps
• Cost Efficiency: Avoid re-running successful operations
• Transparency: .status files show exactly what completed/failed
• Flexible Recovery: Auto-detect restart points or force restart from any step
• Multi-file Support: Independent restart points per input file

Status File Example:

{
  "chain_config": "my_chain.yaml",
  "status": "failed",
  "files": {
    "chapter_21.md": {
      "status": "failed",
      "steps": {
        "1": {"status": "completed", "name": "grammar_check", "time": 136.3},
        "2": {"status": "completed", "name": "style_improvement", "time": 135.7},
        "3": {"status": "failed", "name": "final_polish", "error": "API timeout"}
      }
    }
  }
}

When restarted, the chain will:

1. Detect the previous failure at step 3
2. Skip steps 1-2 (already completed)
3. Resume from step 3 and continue to completion
4. Track all new progress in the status file

Chapter Splitter

Split documents into separate files by chapters, prologues, and epilogues:

# Basic usage - splits in same directory as input
split-chapters book.md

# Works with plain text files too
split-chapters book.txt -o chapters/

# Specify output directory
split-chapters book.md -o chapters/

# Dry run (preview without creating files)
split-chapters book.md --dry-run

# Overwrite existing files (default: skip existing)
split-chapters book.md --overwrite-existing

# Global command usage (after install-global.sh)
split-chapters --help

Features:

• Multiple Formats: Supports both markdown headers and plain text chapter markers
• Prologue and Epilogue: Detects and extracts Prologue/Prolog and Epilogue/Epilog sections
• Roman Numerals: Handles Roman numeral chapter numbers (I, II, III, IV, etc.)
• Smart Naming: Creates files as prologue.md, chapter_1.md, epilogue.md, etc.
• Title Stripping: Removes text after chapter number (e.g., "Chapter 1: The Beginning" becomes chapter_1.md)
• Duplicate Handling: Adds A, B, C suffixes for duplicate sections (chapter_1_A.md, prologue_A.md)
• Preamble Support: Saves content before first section to preamble.md
• Case Insensitive: Detects "chapter", "Chapter", "CHAPTER", etc.

Supported Chapter Formats:

# Markdown header formats:
# Prologue
# Chapter 1
# Chapter I: The Beginning
## Chapter 2 - The Middle
# Epilogue

# Plain text formats (no # prefix):
Prologue
Chapter 1
Chapter I: The Beginning
Chapter 2 - The Middle
Epilogue

Example Output:

Found Prologue at line 5 → prologue.md
  → Wrote 42 lines to prologue.md
Found Chapter 1 at line 50 → chapter_1.md
  → Wrote 1,245 lines to chapter_1.md
Found Chapter 2 at line 1,300 → chapter_2.md
  → Wrote 1,089 lines to chapter_2.md
Found Chapter 3 at line 2,400 → chapter_3.md
  → Wrote 1,367 lines to chapter_3.md
...
Found Chapter 15 at line 15,800 → chapter_15.md
  → Wrote 1,156 lines to chapter_15.md
Found Epilogue at line 17,000 → epilogue.md
  → Wrote 85 lines to epilogue.md

Writing preamble (342 lines) → preamble.md

============================================================
SUMMARY
============================================================
Input file: book.md
Output directory: /path/to/chapters/
Chapters written: 17

Chapter files written:
  Prologue  → prologue.md          (  42 lines)
  Chapter 1 → chapter_1.md         (1,245 lines)
  Chapter 2 → chapter_2.md         (1,089 lines)
  ...
  Chapter 15 → chapter_15.md       (1,156 lines)
  Epilogue  → epilogue.md          (  85 lines)

Use Cases:

• Book Processing: Split manuscripts for chapter-by-chapter editing (markdown or plain text)
• Chain Preparation: Prepare individual chapters for prompt chain processing
• Parallel Processing: Enable concurrent processing of multiple chapters
• Classic Literature: Process books with Roman numeral chapter numbering
• Plain Text Manuscripts: Work with unformatted text files without markdown
• Story Structure: Extract prologues, chapters, and epilogues separately
• Organization: Structure large documents for better version control

Integration with Chains:

# 1. Split book into chapters
split-chapters manuscript.md -o chapters/

# 2. Process each chapter with a chain
for chapter in chapters/chapter_*.md; do
  openrouter-chain -c editing_chain.yaml --input "$chapter"
done

# 3. Combine results if needed
cat chapters/processed_*.md > final_manuscript.md

---

Configuration

Model Support

Supports 100+ models through OpenRouter:

• Anthropic: Claude 4, Claude 3.5 Sonnet, Claude Haiku
• OpenAI: GPT-4 Turbo, GPT-4o, GPT-3.5 Turbo
• Google: Gemini 2.0 Flash, Gemini 1.5 Pro
• Specialized: DeepSeek Coder, Llama models
• And many more...

Advanced Features

Pre/Post Processing Scripts

Execute custom scripts before and after processing:

# Global scripts (run once per chain)
preprocessing:
  script01: echo "Starting processing pipeline"
  name01: "Initialize Pipeline"
  script02: python3 scripts/validate_input.py
  name02: "Validate Input"

postprocessing:
  script01: python3 scripts/generate_report.py
  name01: "Generate Report"

# Per-step scripts (run for each prompt)
prompts:
  prompt 1:
    prescript: "touch {input_file}.backup"
    name: "content_processing"
    prompt_file: "prompts/process.json"
    postscript: "wc -l {output_file} > {output_file}.stats"

Multi-Pass Processing

Run prompts multiple times with iterative improvement:

prompts:
  prompt 1:
    name: "iterative_improvement"
    prompt_file: "prompts/improve.json"
    passes: 3  # Run 3 times
    append: yes  # Accumulate content

File Conversion

Convert between markdown and various document formats:

# Convert input docx to markdown before processing
input_convert:
  enabled: true
  from_format: docx  # Auto-detects if not specified

# Convert final output to multiple formats
output_convert:
  enabled: true
  formats:
    - docx
    - pdf
    - epub

# Per-step conversion for specific outputs
prompts:
  prompt 2:
    convert_output:
      format: pdf
      filename: step2_report.pdf

Supported Formats:

• Input: docx, odt, rtf, html, epub, txt, pdf, latex, rst, org, mediawiki
• Output: docx, odt, rtf, html, epub, pdf, txt, latex, rst, org, mediawiki

Requirements:

• pypandoc Python package (auto-installed)
• pandoc system binary (auto-installed by install scripts)
• LaTeX for PDF generation (optional)

Combine Outputs

Merge all step outputs into a single file:

# Enable combine feature
combine: true
combined_file_name: all_steps_combined.md

prompts:
  prompt 1:
    name: "analysis"
    prompt_file: "prompts/analyze.json"
  prompt 2:
    name: "enhancement"
    prompt_file: "prompts/enhance.json"
  prompt 3:
    name: "finalization"
    prompt_file: "prompts/finalize.json"

# Result: Creates markdown file with headers for each step

Output Format:

# Combined Output - input_file.md
Generated: 2025-01-15 14:30:00
---
## Step 1: analysis
[output from step 1]
---
## Step 2: enhancement
[output from step 2]
---
## Step 3: finalization
[output from step 3]
---

File Management

• Automatic Chunking: Handle large files automatically
• Intermediate Files: All step outputs preserved
• Size Validation: Detect processing issues
• Format Support: Markdown, text, JSON, YAML

---

Examples

Content Enhancement Pipeline

# content_enhancement.yaml
global_config:
  model: "anthropic/claude-4-sonnet-20250522"
  temperature: 0.7

input_file: blog_draft.md
output_file: enhanced_blog.md

prompts:
  prompt 1:
    name: "grammar_foundation"
    prompt_file: "prompts/grammar.json"
    temperature: 0.2

  prompt 2:
    name: "engagement_boost"
    prompt_file: "prompts/engagement.json"
    temperature: 0.8

  prompt 3:
    name: "seo_optimization"
    prompt_file: "prompts/seo.json"
    temperature: 0.6

Technical Documentation

# technical_docs.yaml
global_config:
  model: "deepseek/deepseek-coder"
  temperature: 0.3

preprocessing:
  script01: python3 scripts/extract_code.py
  name01: "Extract Code Blocks"

prompts:
  prompt 1:
    prescript: "python3 scripts/validate_syntax.py {input_file}"
    name: "code_documentation"
    prompt_file: "prompts/document_code.json"
    postscript: "python3 scripts/test_examples.py {output_file}"

postprocessing:
  script01: python3 scripts/generate_toc.py
  name01: "Generate Table of Contents"

Multi-File Processing

# multi_file_processing.yaml
input_files:
  - chapter1.md
  - chapter2.md
  - chapter3.md
output_pattern: "processed_{input_name}.md"

prompts:
  prompt 1:
    prescript: "echo 'Processing {input_file}' >> progress.log"
    name: "content_enhancement"
    prompt_file: "prompts/enhance.json"
    postscript: "wc -w {output_file} >> word_counts.log"

Document Conversion Pipeline

Process Word documents and export to multiple formats:

# document_conversion.yaml
input_file: manuscript.docx
output_file: processed_manuscript.md

# Convert input Word doc to markdown
input_convert:
  enabled: true
  from_format: docx

# Convert final output to multiple formats
output_convert:
  enabled: true
  formats:
    - docx  # Microsoft Word
    - pdf   # PDF document
    - epub  # eBook format

# Combine all steps for reference
combine: true
combined_file_name: processing_history.md

global_config:
  model: "anthropic/claude-4-sonnet-20250522"
  temperature: 0.7

prompts:
  prompt 1:
    name: "grammar_check"
    prompt_file: "prompts/grammar.json"
    temperature: 0.3

  prompt 2:
    name: "style_improvement"
    prompt_file: "prompts/style.json"
    temperature: 0.7
    # Export this step as PDF report
    convert_output:
      format: pdf
      filename: style_report.pdf

  prompt 3:
    name: "final_polish"
    prompt_file: "prompts/polish.json"
    temperature: 0.5

# Result: Creates processed_manuscript.md, .docx, .pdf, .epub
#         Plus processing_history.md with all steps combined
#         Plus style_report.pdf from step 2

Chapter-Based Book Processing

Process a book manuscript chapter by chapter:

# Step 1: Split manuscript into chapters
split-chapters manuscript.md -o manuscript_chapters/

# Step 2: Create a chapter processing chain
cat > chapter_editing.yaml << 'EOF'
global_config:
  model: "anthropic/claude-4-sonnet-20250522"
  temperature: 0.7

input_files:
  - manuscript_chapters/chapter_1.md
  - manuscript_chapters/chapter_2.md
  - manuscript_chapters/chapter_3.md
  # Add all chapters...

output_pattern: "edited_{input_name}.md"

prompts:
  prompt 1:
    name: "grammar_and_clarity"
    prompt_file: "prompts/grammar.json"
    temperature: 0.3

  prompt 2:
    name: "dialogue_enhancement"
    prompt_file: "prompts/dialogue.json"
    temperature: 0.7

  prompt 3:
    name: "pacing_review"
    prompt_file: "prompts/pacing.json"
    temperature: 0.6
    passes: 2
EOF

# Step 3: Process all chapters
openrouter-chain -c chapter_editing.yaml

# Step 4: Combine edited chapters
cat manuscript_chapters/edited_chapter_*.md > final_manuscript.md

# Optional: Convert to multiple formats
openrouter-chain -c convert_final.yaml

With restart functionality:

# If processing fails on chapter 5
openrouter-chain -c chapter_editing.yaml --restart

# Check status of all chapters
openrouter-chain -c chapter_editing.yaml --status-only

# Restart specific chapter from step 2
openrouter-chain -c chapter_editing.yaml --restart-from 2

Web Interface Load Examples

Load Prompt Example

// test_load_prompt.json
{
  "title": "Content Quality Analysis",
  "description": "Analyze content for clarity, coherence, and quality",
  "persona": "You are a professional content analyst who provides clear, actionable feedback.",
  "instructions": "Analyze the provided content for clarity, coherence, and overall quality. Provide specific recommendations for improvement.",
  "review_criteria": "Evaluate based on: 1) Clarity of message, 2) Logical structure, 3) Grammar and style, 4) Overall effectiveness",
  "output_format": "Provide a structured analysis with specific examples and recommendations."
}

Web Usage:

1. Click "Load Prompt" on main page
2. Upload test_load_prompt.json
3. Optionally upload test_single_prompt_config.yaml for custom API settings
4. Enter content: "This is my draft article about AI development..."
5. Execute and view results instantly

Load Chain Example

# test_web_yaml_config.yaml
global_config:
  model: "anthropic/claude-4-sonnet-20250522"
  temperature: 0.7
  max_tokens: 20000

input_file: test_input.md
output_file: test_web_output.md

preprocessing:
  script01: echo "Starting web-loaded chain"
  name01: "Initialize Chain"

prompts:
  prompt 1:
    name: "test_analysis"
    prompt_file: "prompts/content_quality.json"
    temperature: 0.3

  prompt 2:
    name: "final_polish"
    prompt_file: "prompts/content_quality.json"
    temperature: 0.8
    append: yes

postprocessing:
  script01: echo "Web chain complete"
  name01: "Finalize Chain"

Web Usage:

1. Navigate to /chains page
2. Click "Load Chain" button
3. Upload test_web_yaml_config.yaml
4. Preview shows: 2 prompts, preprocessing/postprocessing scripts, append mode
5. Enter content and start execution
6. Monitor real-time progress in active chains section

---

Documentation

Quick References

• Complete Documentation - Full guides and API reference
• Configuration Guide - YAML configuration options
• Prompt Templates - Creating reusable prompts
• Troubleshooting - Common issues and solutions

Common Commands

# Help and information
openrouter-runner --help
openrouter-chain --help
openrouter-web --help
split-chapters --help

# Chapter splitting (supports markdown, plain text, Roman numerals)
split-chapters book.md
split-chapters book.txt -o chapters/
split-chapters book.md --dry-run
split-chapters book.md --overwrite-existing

# Debug and validation
openrouter-chain -c config.yaml --debug
python3 -c "import yaml; print(yaml.safe_load(open('config.yaml')))"

# Model and API testing
openrouter-runner -p prompts/test.json -i test.md -o test_output.md

---

Development

Project Structure

openrouter-interface/
├── src/openrouter_interface/     # Main Python package
│   ├── cli.py                    # CLI interface
│   ├── web.py                    # Web interface
│   ├── chain.py                  # Chain runner
│   ├── prompt_runner.py          # Core processing engine
│   ├── prompt_chain_runner.py    # Chain execution logic
│   └── split_chapters.py         # Chapter splitter utility
├── config/                       # Configuration files
├── prompts/                      # Prompt templates
├── scripts/                      # Utility scripts
├── tests/                        # Test suite
└── docs/                         # Documentation

Running Tests

# Run all tests
pytest

# Run with coverage
pytest --cov=openrouter_interface

# Run specific test files
pytest tests/test_cli.py
pytest tests/test_chain.py

Code Quality

# Format code
black src tests

# Type checking
mypy src

# Linting
flake8 src

---

Contributing

We welcome contributions! Here's how to get started:

Quick Contribution Guide

1. Fork the repository
2. Create a feature branch: git checkout -b feature/amazing-feature
3. Make your changes with tests and documentation
4. Run tests: pytest && black src tests && flake8 src
5. Submit a pull request

Areas for Contribution

• New Features: Prompt templates, chain configurations
• Model Support: Additional model integrations
• Documentation: Tutorials, examples, guides
• Testing: Test coverage improvements
• Performance: Optimization and profiling
• UI/UX: Web interface improvements

Development Setup

# Clone and install for development
git clone https://github.com/your-org/openrouter-interface.git
cd openrouter-interface
./install.sh dev

# Install pre-commit hooks
pip install pre-commit
pre-commit install

# Run development server
PYTHONPATH=src python3 -m openrouter_interface.web --debug

---

Support

Getting Help

• Bug Reports: GitHub Issues
• Discussions: GitHub Discussions
• Documentation: Complete Docs
• Feature Requests: GitHub Issues

Troubleshooting

Common Issues:

# API key not set
echo $OPENROUTER_API_KEY

# Permission issues
chmod +x install-global.sh

# Python path issues
which python3
python3 --version

Debug Mode:

# Enable detailed logging
openrouter-chain -c config.yaml --debug 2>&1 | tee debug.log

# Check configuration validity
python3 -c "import yaml; yaml.safe_load(open('config.yaml'))"

---

License

This project is licensed under the MIT License - see the LICENSE file for details.

---

Acknowledgments

• OpenRouter for providing access to multiple AI models
• The Python community for excellent tools and libraries
• Contributors who help improve this project

---

Project Status

• Stable: Core functionality tested and reliable
• Active Development: Regular updates and new features
• Well Tested: Comprehensive test suite
• Documented: Complete documentation and examples

Recent Updates

• Per-Phase Scripts: Individual prescript/postscript for each step
• Variable Substitution: Dynamic {input_file}/{output_file} replacement
• Enhanced Monitoring: Improved web interface with real-time progress
• Model Compatibility: Automatic parameter filtering per model
• File Management: Advanced validation and size checking
• Load Buttons: Direct upload and execution of JSON prompts and YAML chains
• No-Registry Execution: Run files instantly without permanent storage
• Expanded Load Chain: Full-page configuration interface with rich preview and validation
• YAML Editor: Professional code editor with templates, syntax highlighting, and real-time validation
• Enhanced Error Handling: Comprehensive YAML parsing error detection and reporting
• Chapter Splitter: Standalone utility to split documents by chapters, prologues, and epilogues (supports markdown, plain text, Roman numerals)

---

Ready to get started? Follow the Quick Start guide above!