persona-toolkit 0.1.7

Toolkit for Persona, an agent AI system — provides modular, callable tools for dynamic function execution

Persona CLI

Persona CLI is a command-line tool for creating and managing projects that use the persona-toolkit — a modular function-calling framework designed for agent-based systems.

This CLI helps you scaffold projects, generate tools, test them locally, and run the FastAPI server that exposes your tools via REST API.

---

🚀 Installation

To use the CLI, first install the persona-toolkit library (assuming it's published or available locally):

pip install persona-toolkit

Or if you're developing locally:

cd persona-toolkit/
poetry install

The CLI is exposed as:

persona-toolkit

---

📆 Features

init

Create a new project scaffold with Poetry and the required structure:

persona-toolkit init my-project

This will:

• Initialize a Poetry project
• Install persona-toolkit as a dependency
• Create a tools/ folder where your tools will live

---

add-tool

Add a new tool interactively:

persona-toolkit add-tool

You'll be prompted for a tool name. A new Python file will be created in the tools/ directory with a ready-to-edit template including:

• Input and Output models (using Pydantic)
• A run() function

---

test-tool

Test a tool locally by manually entering its input values:

persona-toolkit test-tool echo

This will:

• Import the specified tool from the tools/ directory
• Prompt for input fields
• Run the run() function and show the output

---

run

Start the FastAPI server and expose your tools via HTTP:

persona-toolkit run --port 8000

You can now access:

• GET /tools — list available tools
• GET /tools/{tool}/schema — get tool schema
• POST /tools/{tool}/invocations — run a tool

---

🗂 Project Structure

my-project/
├── pyproject.toml         # Poetry project config
├── tools/
│   └── echo.py            # Example tool

Each tool must define:

• Input (a Pydantic model)
• Output (a Pydantic model)
• run(input: Input) -> Output

---

💡 Example Tool

# tools/echo.py

from pydantic import BaseModel, Field


class Input(BaseModel):
    message: str = Field(description="Message to echo")


class Output(BaseModel):
    message: str


def run(input: Input) -> Output:
    """Echo the message back"""
    return Output(message=f"Echo: {input.message}")

---

✅ Requirements

• Python 3.9+
• Poetry
• Uvicorn (installed automatically)

---

📃 License

MIT License

---

Built for the Persona Agent System 🤖