CLI

Using the Command Line Interface

The Peeps AI Command-Line Interface (CLI) is a powerful tool designed to simplify the process of creating, managing, and running workflows. Whether you're orchestrating simple automations or complex multi-agent systems, the CLI provides a streamlined, efficient way to control your Peeps AI projects.

Prerequisites

Before running a workflow, ensure that:

  • Python 3.10 - 3.12 is installed on your system.

  • Peeps AI is installed via:

    pip install peepsai
  • Your project structure is set up correctly with agents and tasks defined in YAML configuration files.

Project Initialization

To start a new Peeps AI project:

peepsai create group <project_name>

This command generates a project folder with the essential files:

  • agents.yaml - Defines AI agents and their roles.

  • tasks.yaml - Outlines the tasks each agent will handle.

  • group.py & main.py - Core Python files for defining group logic and running workflows.

Basic CLI Commands

Running a Workflow

The primary command to execute your workflow:

peepsai run

This will trigger the workflow as defined in your configuration files. Alternatively, you can run directly with Python:

python src/<project_name>/main.py

Specifying Input Parameters

You can pass dynamic inputs to your agents:

peepsai run --inputs '{"topic": "AI Automation", "deadline": "2024-12-31"}'

This allows for flexible task execution based on real-time data.

Verbose Mode

To enable detailed logging during workflow execution:

peepsai run --verbose

Verbose mode is helpful for debugging complex workflows, as it provides step-by-step logs of agent interactions and task progress.

Managing Workflows

Listing Available Groups

peepsai list groups

Displays all defined groups within your project.

Inspecting Group Details

peepsai inspect group <group_name>

Provides details about agents, tasks, and configurations within a specific group.

Running Specific Tasks

If you want to execute a specific task without running the entire workflow:

peepsai run --task <task_name>

This is useful for testing individual components during development.

Advanced CLI Features

Custom Configuration Files

Override default settings with custom configurations:

peepsai run --config path/to/custom_config.yaml

This enables you to run the same workflow with different agent behaviors or task definitions.

Parallel Execution

Leverage multi-agent parallelism:

peepsai run --parallel

This command allows tasks to run concurrently, reducing execution time for large workflows.

Error Handling and Troubleshooting

  • Common Error: ModuleNotFoundError: No module named 'tiktoken'

    • Fix: Install missing dependencies:

      pip install 'peepsai[embeddings]'
  • Workflow Crashes: Check the logs with:

    peepsai run --verbose

    This provides detailed error traces to identify issues quickly.

Automating Workflow Execution

For continuous deployment or automation pipelines, integrate Peeps AI workflows with cron jobs or CI/CD systems:

0 * * * * /usr/bin/python3 /path/to/project/src/project_name/main.py >> /var/log/peepsai.log 2>&1

This runs the workflow every hour and logs outputs for monitoring.


Last updated