Setup & Installation
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
Version History
Updated to use Python 3.12, EvoAgentX now installed
What This Skill Does
EvoAgentX is a framework for building AI agents that auto-construct workflows from a stated goal and improve themselves based on feedback. It supports multiple LLM backends and includes short-term and long-term memory. The skill provides install, status check, docs, and workflow execution commands via PowerShell.
Auto-constructing workflows from a goal description removes the need to manually wire together agent steps that most frameworks require you to define upfront.
When to use it
- Running a multi-step research task by describing a goal in plain text
- Automating a repetitive workflow that requires iterative refinement
- Testing how different LLM backends handle the same agent goal
- Building a prototype agent pipeline without writing orchestration code manually
- Checking which API keys and dependencies are configured before running a job
Example Workflow
Here's how your AI assistant might use this skill in practice.
User asks: run an EvoAgentX workflow with the goal 'summarize top AI papers from last week'
- 1Sets OPENAI_API_KEY environment variable
- 2Runs evoagentx.ps1 -Action status to confirm installation and key presence
- 3Runs evoagentx.ps1 -Action run -Goal 'summarize top AI papers from last week'
- 4EvoAgentX auto-constructs a workflow, executes agent steps, and applies self-evolution based on intermediate feedback
A summary of recent AI papers produced by the self-constructed agent workflow
Requirements
Accounts, API keys, or tools you or your AI assistant may need to set up while using this skill.