Running Cleo
Cleo is designed to operate in flexible environments — from local terminals to remote servers and containerized deployments. Whether you're experimenting with agent behavior or integrating Cleo into a production system, this guide walks through the key steps to get Cleo running cleanly and consistently.
Modes of Operation
Cleo can be run in several modes depending on your environment and needs:
Interactive Terminal: Default CLI-based interaction loop.
Scripted Execution: Batch tasks fed via code or external scripts.
Daemonized Agent Loop: Background service for persistent agents.
API-Based Operation (coming soon): REST interface for remote communication.
Initial Setup
Ensure Python 3.10+ is installed. Then:
Edit .env
with your desired configuration for memory backend, agent defaults, and other runtime options.
Running Cleo Locally
Method 1: Launch Default Agent
This will load the default agent defined in main.py
and enter an interactive task loop:
Method 2: Load a Specific Agent
Add a CLI interface to support flexible agent selection and task chaining (if not already implemented).
Logging and Debugging
Cleo includes structured logging (in logs/
) and agent-specific transcripts.
You can add runtime debug flags via .env
:
Logging output includes timestamps, task IDs, selected tools, and memory hits.
Subpages
To better support diverse workflows, this section includes three detailed subpages:
1. Local Deployment
For users developing and testing on their own machines.
2. Server & Container Environments
Deploying Cleo to a Linux server, Docker, or cloud instance.
3. Scheduled & Background Execution
Setting up persistent agents and CRON/daemon workflows.
Would you like to begin with the Local Deployment subpage?
4o
You said:
yes
ChatGPT said:
Here is the Local Deployment subpage under the “Running Cleo” section. This page walks users through setting up, running, and testing Cleo on their local machine in a stable and development-ready environment.
Last updated