CLI Commands

Complete reference for all Erdo CLI commands and their options.

Global Options

Available for all commands:

--help, -h: Show help information
--version, -v: Show version information
--config: Specify config file path
--verbose: Enable verbose output

Authentication Commands

Authenticate with Erdo services.

erdo login

Options:

--token: Login with API token
--org: Specify organization

Examples:

# Interactive login
erdo login

# Login with token
erdo login --token your_api_token

`erdo logout`

Sign out of current session.

erdo logout

Project Commands

`erdo init`

Initialize a new Erdo project.

erdo init [project-name]

Options:

--template: Use project template
--python-version: Specify Python version

Examples:

# Create new project
erdo init my-project

# Use template
erdo init my-project --template data-analysis

`erdo dev`

Start development server.

erdo dev

Options:

--port: Specify port (default: 3000)
--host: Specify host (default: localhost)
--watch: Enable file watching

`erdo build`

Build project for deployment.

erdo build

Options:

--output: Output directory
--production: Production build

Agent Commands

`erdo agent create`

Create a new agent.

erdo agent create <name>

Options:

--template: Agent template
--description: Agent description

Examples:

# Create basic agent
erdo agent create data-analyzer

# Use template
erdo agent create sales-bot --template customer-service

`erdo agent list`

List all agents.

erdo agent list

Options:

--status: Filter by status
--tag: Filter by tag

`erdo agent run`

Run an agent.

erdo agent run <agent-name>

Options:

--input: Input data file
--params: Parameters JSON
--async: Run asynchronously

Examples:

# Run agent with input
erdo agent run data-analyzer --input data.csv

# Run with parameters
erdo agent run sales-bot --params '{"customer_id": "123"}'

`erdo agent-test`

Run agent tests with parallel execution.

erdo agent-test <test_file.py>

Description: The agent-test command discovers all agent_test_* functions in your Python test files and runs them in parallel for fast execution. Tests use the invoke() function with mode="replay" for fast, free testing. Options:

-v, --verbose: Show detailed error traces
-j, --jobs <N>: Number of parallel jobs (default: auto)
-r, --refresh: Force refresh cached responses in replay mode

Examples:

# Run all tests in a file
erdo agent-test tests/test_my_agent.py

# Verbose output (show full error traces)
erdo agent-test tests/test_my_agent.py --verbose

# Limit parallel jobs
erdo agent-test tests/test_my_agent.py -j 4

# Refresh cached responses (bypass cache in replay mode)
erdo agent-test tests/test_my_agent.py --refresh

Test Pattern: Write test functions with the agent_test_* prefix:

from erdo import invoke
from erdo.test import text_contains

def agent_test_basic_query():
    """Test basic agent invocation."""
    response = invoke(
        "my-agent",
        messages=[{"role": "user", "content": "Hello"}],
        mode="replay",  # Free after first run!
    )

    assert response.success
    assert text_contains(str(response.result), "expected text")

Output Example:

Discovering tests in tests/test_my_agent.py...
Found 15 tests

Running tests in parallel...

======================================================================
AGENT TEST RESULTS
======================================================================

✅ agent_test_csv_sales_total (0.45s)
✅ agent_test_csv_product_breakdown (0.52s)
❌ agent_test_invalid_dataset (0.21s)

----------------------------------------------------------------------
Total: 15 | Passed: 14 | Failed: 1 | Duration: 2.3s
----------------------------------------------------------------------

See Also:

Integration Testing - Complete integration testing guide
Testing Overview - Learn about testing strategies
Invoke - Learn about the invoke() function

`erdo export-bot`

Export agents for backup, sharing, migration, or local development.

erdo export-bot <agent-name>

Basic Usage:

# Export to stdout (standalone Python file)
erdo export-bot "file analyzer"

# Export to file (standalone Python file)
erdo export-bot "file analyzer" > my-agent.py

# Export with separate code files (-o flag)
erdo export-bot "file analyzer" -o my-agent-with-files.py

Export Formats:

Standalone Export (Default)
Package Structure Export (-o flag)

# Creates a single Python file with all code embedded
erdo export-bot "file analyzer" > standalone-agent.py

Features:

Single file contains complete agent definition
All Python code files are embedded with fixed imports
Ready for syncing back with erdo sync
Common.py classes are moved to the top of the file
Relative imports are commented out with explanatory notes

# Creates main file + separate directory with code files
erdo export-bot "file analyzer" -o agent-with-files.py

Creates:

agent-with-files.py - Main agent definition (syncable)
file_analyzer_analyze_file_files/ - Directory with individual Python files
analyze_file.py - Main analysis code with relative imports
analyze_file_fn.py - Analysis functions
common.py - Shared classes and utilities

Benefits:

Individual files maintain proper Python package structure
Relative imports work correctly in the separate files
Easier for local development and testing
Can be used as a proper Python package

Options:

Option	Description	Example
`-o, --output`	Export with separate code files	`erdo export-bot "my-agent" -o agent.py`
`--format`	Export format (python, json, yaml)	`erdo export-bot "my-agent" --format json`
`--include-deps`	Include dependencies	`erdo export-bot "my-agent" --include-deps`

Examples:

# Export for backup
erdo export-bot "data-processor" > backups/data-processor-$(date +%Y%m%d).py

# Export for local development
erdo export-bot "file-analyzer" -o local-development.py
cd file_analyzer_analyze_file_files/
python analyze_file.py  # Run individual files

# Export for sharing
erdo export-bot "security-checker" --format json > agent-config.json

# Export multiple agents
for agent in "file analyzer" "data analyst" "security checker"; do
  erdo export-bot "$agent" > "exports/${agent// /-}.py"
done

Syncing Exported Agents:

# Sync standalone export back to platform
erdo sync exported-agent.py --name-suffix="_imported"

# Sync package structure export (use main file)
erdo sync agent-with-files.py --name-suffix="_from_package"

# The separate code files directory is for local development only

`erdo sync`

`erdo invoke`

Invoke an agent from the command line.

erdo invoke <agent-name> [options]

Description: Execute an agent with messages and parameters directly from the CLI. Returns structured results following the executor pattern with clean separation of result, messages, steps, and events. Options:

-m, --message <text>: User message to send to the agent
-p, --parameters <json>: Parameters as JSON string
-d, --datasets <list>: Comma-separated dataset slugs
--mode <mode>: Invocation mode (live, replay, manual)
--json: Output full structured JSON result
-v, --verbose: Show detailed steps during execution
--stream: Stream output in real-time

Examples:

# Simple invocation
erdo invoke data-question-answerer -m "What were Q4 sales?"

# With datasets
erdo invoke data-analyzer -m "Analyze trends" -d sales-2024,customers

# With parameters
erdo invoke file-processor -m "Process files" -p '{"format": "csv"}'

# Verbose output (shows steps)
erdo invoke data-question-answerer -m "Hello" --verbose

# JSON output (full structured result)
erdo invoke data-analyzer -m "Analyze" --json

# Using replay mode (cached responses)
erdo invoke my-agent -m "Test" --mode replay

Output Formats: By default, erdo invoke outputs the agent’s text response:

$ erdo invoke data-question-answerer -m "Hello"
Hello! I'm ready to help answer questions about your data.

With --verbose, it shows steps before the output:

$ erdo invoke data-question-answerer -m "Hello" --verbose
Steps:
  ✓ parse_input (utils.parse_json)
  ✓ generate_response (llm.message)
  ✓ format_output (utils.echo)

Hello! I'm ready to help answer questions about your data.

With --json, it returns the complete structured result:

$ erdo invoke data-analyzer -m "Test" --json
{
  "success": true,
  "bot_id": "data-analyzer",
  "invocation_id": "inv_abc123",
  "result": {
    "status": "success",
    "parameters": {},
    "output": {
      "content": [
        {"content_type": "text", "content": "Analysis complete..."}
      ]
    },
    "message": null,
    "error": null
  },
  "messages": [
    {"role": "assistant", "content": "Analysis complete..."}
  ],
  "steps": [
    {"key": "analyze", "action": "llm.message", "status": "completed"}
  ],
  "error": null
}

Result Structure: The invoke command follows the executor pattern with clean separation:

result: The types.Result object with status/parameters/output/message/error
messages: All messages from all steps (including sub-agents)
steps: Information about executed steps
events: Complete raw event stream (only in JSON mode)

See Also:

SDK Invoke - Programmatic invoke() usage
Integration Testing - Testing with invoke

Integration Commands

`erdo integration add`

Add a new integration.

erdo integration add <type>

Options:

--name: Integration name
--config: Configuration file

Examples:

# Add database integration
erdo integration add postgresql --name prod-db

# Add with config file
erdo integration add salesforce --config sf-config.json

`erdo integration list`

List all integrations.

erdo integration list

Options:

--type: Filter by type
--status: Filter by status

`erdo integration test`

Test integration connection.

erdo integration test <integration-name>

`erdo integration remove`

Remove an integration.

erdo integration remove <integration-name>

Options:

--force: Force removal without confirmation

Deployment Commands

`erdo deploy`

Deploy project to Erdo cloud.

erdo deploy

Options:

--env: Target environment
--tag: Deployment tag
--dry-run: Preview deployment

Examples:

# Deploy to production
erdo deploy --env production

# Dry run
erdo deploy --dry-run

`erdo status`

Check deployment status.

erdo status

Options:

--env: Environment to check
--detailed: Show detailed status

Data Commands

`erdo data import`

Import data from external sources.

erdo data import <source>

Options:

--format: Data format (csv, json, etc.)
--destination: Target dataset

Examples:

# Import CSV file
erdo data import data.csv --format csv

# Import to specific dataset
erdo data import data.json --destination my-dataset

`erdo data export`

Export data to external formats.

erdo data export <dataset>

Options:

--format: Export format
--output: Output file path

`erdo data list`

List available datasets.

erdo data list

Options:

--type: Filter by data type
--size: Filter by size

Configuration Commands

`erdo config get`

Get configuration value.

erdo config get <key>

`erdo config set`

Set configuration value.

erdo config set <key> <value>

Examples:

# Set API endpoint
erdo config set api.endpoint https://api.erdo.ai

# Set default model
erdo config set llm.default_model claude-sonnet-4

`erdo config list`

List all configuration values.

erdo config list

Monitoring Commands

`erdo logs`

View application logs.

erdo logs

Options:

--follow: Follow log output
--lines: Number of lines to show
--level: Filter by log level

Examples:

# View recent logs
erdo logs --lines 100

# Follow logs
erdo logs --follow

`erdo metrics`

View performance metrics.

erdo metrics

Options:

--agent: Filter by agent
--time-range: Time range for metrics

Utility Commands

`erdo version`

Show version information.

erdo version

`erdo help`

Show help information.

erdo help [command]

`erdo update`

Update Erdo CLI to latest version.

erdo update

Options:

--pre-release: Include pre-release versions

Environment Variables

Configure Erdo CLI behavior:

ERDO_API_TOKEN: API authentication token
ERDO_API_ENDPOINT: API endpoint URL
ERDO_CONFIG_PATH: Configuration file path
ERDO_LOG_LEVEL: Logging level (debug, info, warn, error)

Configuration File

Default configuration file (~/.erdo/config.yaml):

api:
  endpoint: https://api.erdo.ai
  token: your_api_token

defaults:
  model: claude-sonnet-4
  timeout: 30s

logging:
  level: info
  format: json

Exit Codes

0: Success
1: General error
2: Authentication error
3: Configuration error
4: Network error
5: Validation error

For detailed help on any command, use erdo help <command> or erdo <command> --help.

Get Started

Core Concepts

Testing & Validation

Examples & Use Cases

Python SDK

TypeScript SDK

CLI Tools

Pre-built Agents

Advanced

Integrations

​CLI Commands

​Global Options

​Authentication Commands

​erdo login

​erdo logout

​Project Commands

​erdo init

​erdo dev

​erdo build

​Agent Commands

​erdo agent create

​erdo agent list

​erdo agent run

​erdo agent-test

​erdo export-bot

​erdo sync

​erdo invoke

​Integration Commands

​erdo integration add

​erdo integration list

​erdo integration test

​erdo integration remove

​Deployment Commands

​erdo deploy

​erdo status

​Data Commands

​erdo data import

​erdo data export

​erdo data list

​Configuration Commands

​erdo config get

​erdo config set

​erdo config list

​Monitoring Commands

​erdo logs

​erdo metrics

​Utility Commands

​erdo version

​erdo help

​erdo update

​Environment Variables

​Configuration File

​Exit Codes

CLI Commands

Global Options

Authentication Commands

`erdo login`

`erdo logout`

Project Commands

`erdo init`

`erdo dev`

`erdo build`

Agent Commands

`erdo agent create`

`erdo agent list`

`erdo agent run`

`erdo agent-test`

`erdo export-bot`

`erdo sync`

`erdo invoke`

Integration Commands

`erdo integration add`

`erdo integration list`

`erdo integration test`

`erdo integration remove`

Deployment Commands

`erdo deploy`

`erdo status`

Data Commands

`erdo data import`

`erdo data export`

`erdo data list`

Configuration Commands

`erdo config get`

`erdo config set`

`erdo config list`

Monitoring Commands

`erdo logs`

`erdo metrics`

Utility Commands

`erdo version`

`erdo help`

`erdo update`

Environment Variables

Configuration File

Exit Codes