Use AIGNE CLI to run your agents.

aigne run Command#

Run an agent in the current directory or a specified directory.

Options#

• --path <path>: Path to the agent directory (defaults to current directory .) or URL pointing to a project in AIGNE Studio
• --entry-agent <agent>: Name of the agent to run (defaults to the first agent found)
• --cache-dir <directory>: Directory for downloaded packages (when using URL), defaults to ~/.aigne/xxx
• --chat: Run chat loop in terminal (defaults to false, i.e., one-shot mode)
• --input -i <input>: Input for the agent
• --model <provider[:model]>: AI model to use, format 'provider[:model]' where model is optional. Examples: 'openai' or 'openai:gpt-4o-mini'. Available providers: openai, anthropic, bedrock, deepseek, gemini, ollama, openrouter, xai (defaults to openai)
• --temperature <temperature>: Model temperature parameter (controls randomness, higher values produce more random output). Range: 0.0-2.0, 0.0 for most deterministic output.
• --top-p <top-p>: Model Top P (nucleus sampling) parameter (controls diversity). Range: 0.0-1.0, lower values restrict to more likely tokens.
• --presence-penalty <presence-penalty>: Model presence penalty parameter (penalizes repeated tokens). Range: -2.0 to 2.0, positive values reduce repetition.
• --frequency-penalty <frequency-penalty>: Model frequency penalty parameter (penalizes token usage frequency). Range: -2.0 to 2.0, positive values reduce repeated token usage.
• --log-level <log-level>: Verbose log level for debugging information. Options: "debug", "info", "warn", "error".
• --help: Show command help

Basic Usage Examples#

Run agent in current directory#

Run agent in specific directory#

Run specific agent#

Run agent in chat mode#

Run agent with specific input#

Running Agents from AIGNE Studio#

You can run agents directly from AIGNE Studio project URLs:

Complete Steps to Run Agents from AIGNE Studio#

https://github.com/user-attachments/assets/f528d1a1-31d1-48e5-b89e-e3d555c53649

1. Create a project in AIGNE Studio:
2. • Log in to AIGNE Studio
   • Click to create a new project
   • Add required Agents to the project
2. Get CLI command:
3. • In the project page, click the settings icon in the top right corner
   • Find the "Integration" tab in the settings menu
   • In the AIGNE CLI section, click "Generate Link" button if no link exists
   • Copy the generated aigne run command (format like aigne run https://www.aigne.io/projects/xxx/xxx.tgz?secret=yyy&hash=zzz)
3. Run the command:
4. • Open terminal
   • Paste and run the copied command
   • The system will automatically download the project and its agents and start the chat loop

Model Configuration Examples#

Using different AI models#

Adjusting model parameters#

Debugging and Logging#

Use different log levels to get more debugging information:

debug logging
aigne run --log-level debug

# Show only error messages
aigne run --log-level error