Agent CLI

agent-cli is a Rust library (with an optional CLI) implementing the Agent Life Cycle framework, inspired by the sandy-mount/sandymount Wiki. It provides a modular and reusable implementation for managing intelligent, autonomous software agents through their lifecycle phases: Create, Add Skills, Configure, Run, and Shutdown.

Features

• Library Crate: Core logic in src/lib.rs for creating and managing agents through lifecycle phases.

• Optional CLI: Binary in src/main.rs (if included) allows running commands like agent create my-agent.

• Lifecycle Phases:

  • Create: Initialize an agent with a unique ID and name.
  • Add Skills: Add capabilities (e.g., "memory", "web-access") to the agent.
  • Configure: Set key-value configurations (e.g., goals, permissions).
  • Run: Start the agent, displaying its skills and configuration.
  • Shutdown: Save agent state to a JSON file and decommission the agent.

• Testing: Comprehensive unit tests for all lifecycle phases.

Installation

Prerequisites

• Rust (latest stable version, e.g., via rustup)
• Cargo (included with Rust)

Setup

Clone or create the project:

git clone https://github.com/AI-Robotic-Labs/ai-agent-cli.git
cd agent-cli

OR create a new project:

cargo new agent-cli
cd agent-cli

Ensure your Cargo.toml includes:

[package]
name = "agent-cli"
version = "0.1.0"
edition = "2021"

[dependencies]
serde = { version = "1.0", features = ["derive"] }
serde_json = "1.0"
rand = "0.8"

• Place the library code in src/lib.rs (see Project Structure).
• (Optional) Add src/main.rs for CLI functionality.

Build

cargo build

Usage

As a Library

The core functionality is in src/lib.rs, which defines the AgentCLI struct and its lifecycle methods. Use it in another Rust project by adding agent-cli as a dependency in your Cargo.toml:

[dependencies]
agent-cli = { path = "/path/to/agent-cli" }

Example usage:

use agent_cli::AgentCLI;

fn main() {
    let mut cli = AgentCLI::new();
    cli.create("my-agent").unwrap();
    cli.add_skill("memory").unwrap();
    cli.configure("goal", "assist user").unwrap();
    cli.run().unwrap();
    cli.shutdown().unwrap();
}

As a CLI (Optional)

If src/main.rs is included, you can run commands like:

cargo run -- create my-agent
cargo run -- skill add memory
cargo run -- skill add web-access
cargo run -- config goal "assist user"
cargo run -- run
cargo run -- shutdown

Example Output:

Created agent: my-agent (ID: agent-<random_number>)
Added skill 'memory' to agent my-agent
Added skill 'web-access' to agent my-agent
Configured goal = assist user for agent my-agent
Running agent my-agent (ID: agent-<random_number>)
Active skills: ["memory", "web-access"]
Configuration: {"goal": "assist user"}
Agent shutdown and state saved

The shutdown command saves the agent state to a JSON file (e.g., agent-<random_number>.json).

Testing

Run unit tests in src/lib.rs:

cargo test

Example Output:

running 5 tests
test tests::test_create_agent ... ok
test tests::test_add_skill ... ok
test tests::test_configure ... ok
test tests::test_run ... ok
test tests::test_shutdown ... ok

test result: ok. 5 passed; 0 failed

Project Structure

agent-cli/
├── Cargo.toml
└── src/
    ├── lib.rs   # Core library with AgentCLI and lifecycle logic
    └── main.rs  # Optional CLI binary (add for CLI usage)

Contributing

• Report issues or suggest features via GitHub Issues.
• Submit pull requests with improvements or additional lifecycle phases (e.g., Scale, Monitor).

License

MIT License. See LICENSE for details.

Acknowledgments

• Inspired by the Agent Life Cycle from sandy-mount/sandymount.
• Built with Rust for modularity and safety.