Agent CLI

Command-line interface for single Agent chat and run operations

Agent CLI

The PraisonAI TypeScript CLI provides commands for chatting with and running agents directly from the command line.

Commands Overview

# Chat with an agent
praisonai-ts chat "Your message here"

# Run an agent with a task
praisonai-ts run "Your task" --agent my-agent.yaml

chat

Interactive chat with an agent.

Basic Usage

praisonai-ts chat "Hello, how are you?"

With Model Selection

praisonai-ts chat "Hello" --model gpt-4o-mini

With Streaming

praisonai-ts chat "Write a poem" --stream

Options

Option	Description	Default
--model, -m	LLM model to use	gpt-4o-mini
--stream, -s	Enable streaming output	false
--session	Session ID for conversation continuity	Auto-generated
--verbose, -v	Enable verbose output	false
--json	Output as JSON	false

Examples

# Simple chat
praisonai-ts chat "What is the capital of France?"

# With custom model
praisonai-ts chat "Explain quantum computing" --model gpt-4o

# With session persistence
praisonai-ts chat "My name is Alice" --session user-123
praisonai-ts chat "What's my name?" --session user-123

# JSON output
praisonai-ts chat "Hello" --json

# Streaming output
praisonai-ts chat "Write a story about a robot" --stream

JSON Output Format

{
  "success": true,
  "data": {
    "response": "Hello! How can I help you today?",
    "sessionId": "550e8400-e29b-41d4-a716-446655440000"
  },
  "meta": {
    "duration_ms": 1234,
    "model": "gpt-4o-mini"
  }
}

run

Run an agent with a specific task.

Basic Usage

praisonai-ts run "Analyze this data" --agent my-agent.yaml

Options

Option	Description	Default
--agent, -a	Agent configuration file or name	-
--tools, -t	Comma-separated list of tools	-
--verbose, -v	Enable verbose output	false
--json	Output as JSON	false

Examples

# Run a task with tools
praisonai-ts run "Search the web for AI news" --tools tavily

# Run with agent config
praisonai-ts run "Summarize the document" --agent researcher.yaml

# JSON output for scripting
praisonai-ts run "Generate 5 random names" --json

Environment Variables

Set these environment variables for API access:

# Required - at least one provider key
export OPENAI_API_KEY=sk-...
export ANTHROPIC_API_KEY=sk-ant-...
export GOOGLE_API_KEY=...

# Optional - default model
export PRAISONAI_MODEL=gpt-4o-mini

# Optional - default database for persistence
export PRAISONAI_DB=sqlite:./data.db

Database Persistence

Use the --db flag to persist memory, sessions, and knowledge across CLI invocations.

Supported Databases

Database	URL Format
SQLite	sqlite:./data.db
PostgreSQL	postgres://user:pass@host:5432/db
Redis	redis://localhost:6379

Memory Persistence

# Add memory that persists
praisonai-ts memory add "User prefers dark mode" --db sqlite:./data.db

# List persisted memories
praisonai-ts memory list --db sqlite:./data.db

# Search memories
praisonai-ts memory search "dark" --db sqlite:./data.db

Session Persistence

# Create a persistent session
praisonai-ts session create my-session --db sqlite:./data.db

# List all sessions
praisonai-ts session list --db sqlite:./data.db

# Get session details
praisonai-ts session get my-session --db sqlite:./data.db

Knowledge Persistence

# Add knowledge from text
praisonai-ts knowledge add "PraisonAI is an AI agent framework" --db sqlite:./data.db

# Add knowledge from file
praisonai-ts knowledge add ./document.txt --db sqlite:./data.db

# Search knowledge base
praisonai-ts knowledge search "agent framework" --db sqlite:./data.db

# List all knowledge entries
praisonai-ts knowledge list --db sqlite:./data.db

Environment Variable

Instead of passing --db on every command, set the environment variable:

export PRAISONAI_DB=sqlite:./data.db

# Now all commands use persistent storage
praisonai-ts memory add "Hello world"
praisonai-ts memory list  # Shows persisted entry

Scripting Examples

Bash Script

#!/bin/bash

# Run agent and capture output
RESULT=$(praisonai-ts chat "What is 2+2?" --json)
ANSWER=$(echo $RESULT | jq -r '.data.response')
echo "Answer: $ANSWER"

Pipeline Usage

# Chain with other commands
praisonai-ts chat "Generate a topic" | \
  xargs -I {} praisonai-ts chat "Write about: {}"

Error Handling

Missing API Key

$ praisonai-ts chat "Hello"
Error: OPENAI_API_KEY environment variable is not set

Invalid Model

$ praisonai-ts chat "Hello" --model invalid-model
Error: Model 'invalid-model' is not available

See Also

• Agent - Agent class documentation
• Agents CLI - Multi-agent CLI commands
• Workflow CLI - Workflow CLI commands