- Step1: Install via pip install semi-agent
- Step2: Import Agent and Tool classes in your Python script
- Step3: Define your custom tools with execute methods
- Step4: Initialize the Agent with an LLM, memory store, and tools
- Step5: Provide prompts or goals to the agent instance
- Step6: Call agent.run() to start planning and execution
- Step7: Inspect agent logs and memory to refine workflows