oclawma 0.4.1

OpenClaw Workflow Management Agent

OCLAWMA

OpenClaw Workflow Management Agent

A powerful, extensible AI agent framework with lazy-loaded skills, multi-provider LLM support, and intelligent context management.

🚀 Quick Start

Installation

# Install from PyPI
pip install oclawma

# Or install with development dependencies
pip install oclawma[dev]

Start an Interactive Session

# Start with default local model (requires Ollama)
oclawma run

# Use a specific model
oclawma run --model llama3.2:latest

# Use Kimi cloud provider
export KIMI_API_KEY="your-api-key"
oclawma run --provider kimi --model k2.5

# Auto provider (local with cloud fallback)
oclawma run --provider auto

First Conversation

╔══════════════════════════════════════════════════════════╗
║           🦀 Welcome to OCLAWMA v0.2.0                   ║
║       OpenClaw Workflow Management Agent                 ║
╚══════════════════════════════════════════════════════════╝

You: Hello! Can you help me with my project?

Assistant: Hello! I'd be happy to help you with your project. What would you like to work on?

You: exit

✨ Features

• 🧠 Multi-Provider LLM Support: Ollama (local), Kimi (cloud), and auto-fallback
• 🔧 Lazy-Loaded Skills: Skills load on-demand for fast startup
• 📦 Pip-Installable Skills: Distribute skills via PyPI
• 💬 Interactive Sessions: REPL-style conversation with tool calling
• 📊 Context Budget: Token usage tracking with proactive management
• 🗜️ Context Compression: Automatic summarization when approaching limits
• 🛡️ Safety Controls: Configurable safety levels for tool execution
• 🔌 Extensible: Easy to add new providers, tools, and skills

📖 Documentation

• Installation Guide - Detailed installation instructions
• Configuration Reference - Environment variables and settings
• Skill Development Guide - Create your own skills
• Architecture Overview - System design and components
• API Reference - Python API documentation

🛠️ Built-in Tools

Tool	Description
read	Read file contents
write	Write content to files
exec	Execute shell commands

🧩 Available Skills

Install additional skills via pip:

# Docker skill
pip install oclawma-skill-docker

# Kubernetes skill
pip install oclawma-skill-kubernetes

# GitHub skill
pip install oclawma-skill-github

💻 Interactive Commands

During a session, use these commands:

Command	Description
exit, quit	End the session
/clear	Clear conversation history
/compact	Summarize conversation to save tokens
/status	Show session status and token budget
/help	Show available commands

🏗️ Architecture

┌─────────────────────────────────────────────────────────────┐
│                        CLI Layer                            │
│                    (Click commands)                         │
└─────────────────────────────────────────────────────────────┘
                              │
        ┌─────────────────────┼─────────────────────┐
        ▼                     ▼                     ▼
┌──────────────┐    ┌─────────────────┐    ┌──────────────┐
│   Session    │    │    Providers    │    │    Skills    │
│   Runner     │◄──►│  (Ollama/Kimi)  │    │   Registry   │
└──────────────┘    └─────────────────┘    └──────────────┘
        │                                           │
        ▼                                           ▼
┌──────────────┐                         ┌─────────────────┐
│Context Budget│                         │  Lazy Loading   │
│  Tracker     │                         │   System        │
└──────────────┘                         └─────────────────┘

🤝 Contributing

We welcome contributions! Please see CONTRIBUTING.md for guidelines.

📄 License

MIT License - see LICENSE file for details.

🔗 Links

• Documentation
• PyPI
• Issues
• Discussions