ctxtoolkit 0.1.1

Context Engineering Toolkit - Practical tools for optimizing AI context management

Context Engineering Toolkit (ctxtoolkit)

English | 简体中文

A practical toolkit for optimizing AI context management, helping solve problems such as content loss in long contexts, insufficient tokens, information redundancy, and context pollution.

Core Features

Context Builder

• Key information prioritization optimization
• Intelligent scene background integration
• Structured content layering

Token Saver

• Automatic duplicate content merging
• Terminology compression
• Content summary generation

Anti-Pollution System

• Error information isolation
• Term consistency checking
• Clear task boundary definition

Tool Coordinator

• Tool boundary definition
• Dynamic call constraints
• Multi-tool collaboration workflow

📦 Installation

pip install ctxtoolkit

🚀 Quick Start

1. Context Building

from ctxtoolkit import ContextBuilder

# Create context builder
builder = ContextBuilder()

# Add core instruction
builder.add_core_instruction(
    "Optimize the performance of this Python code",
    requirements=[
        "Reduce memory usage",
        "Improve execution speed",
        "Maintain original functionality"
    ]
)

# Add key information
builder.add_key_info(
    "Code functionality", "Processes 1 million user logs"
)
builder.add_key_info(
    "Current bottleneck", "Nested loops causing O(n²) complexity"
)
builder.add_key_info(
    "Available resources", "8GB RAM, 4-core CPU"
)

# Add supplementary reference
current_code = """
def process_logs(logs):
    results = []
    for i in range(len(logs)):
        for j in range(i+1, len(logs)):
            if logs[i]['user_id'] == logs[j]['user_id']:
                results.append((logs[i], logs[j]))
    return results
"""
builder.add_reference(current_code)

# Generate optimized context
optimized_context = builder.build()
print(optimized_context)

2. Token Saving Example

from ctxtoolkit import TokenSaver

# Create Token saver
saver = TokenSaver()

# Define terminology
saver.add_terminology("R1", "Input format: JSON object containing name(str), age(int), tags(list[str])")
saver.add_terminology("R2", "Output format: Markdown table containing user information and tag statistics")
saver.add_terminology("R3", "Processing rules: Filter age>18, sort by number of tags descending")

# Process user data
user_data = [
    '{"name":"Zhang San","age":25,"tags":["Python","AI"]}',
    '{"name":"Li Si","age":17,"tags":["Java"]}'
]

# Generate compact context
compact_context = saver.build_compact_context(
    "Please process the following user data",
    data=user_data,
    rules=["R1", "R2", "R3"]
)
print(compact_context)

📁 Project Structure

ctxtoolkit/
├── ctxtoolkit/               # Core package directory
│   ├── __init__.py          # Package initialization file
│   ├── context_builder.py   # Context builder
│   ├── token_saver.py       # Token saver
│   ├── anti_pollution.py    # Anti-pollution system
│   └── tool_coordinator.py  # Tool coordinator
├── LICENSE
├── MANIFEST.in
├── README.md
└── setup.py

📚 API Documentation

API documentation has been moved to API_DOCUMENTATION.md.

🔧 Development

Install Development Dependencies

pip install -e .[dev]

Run Tests

pytest

Code Style Check

flake8

🤝 Contributing

Contributions are welcome! Feel free to submit Issues and Pull Requests.

📄 License

MIT License