Pydantic AI Todo

Task planning and tracking with subtasks and dependencies

Hierarchical task management toolset for AI agents with subtask support, dependency tracking, cycle detection, and multiple storage backends (in-memory, PostgreSQL).

Installation

pip install pydantic-ai-todo

GitHub PyPI

Give your AI agents the ability to break down complex tasks into manageable subtasks with dependency tracking. The toolset provides 8 tools including add_todo, add_subtask, set_dependency, complete_todo, and get_available_tasks. Automatic cycle detection prevents circular dependencies. Choose between in-memory storage for quick sessions or PostgreSQL for persistent, multi-tenant task management with session-based isolation.

Features

— Subtask Hierarchy

— Dependency Tracking

— Cycle Detection

— PostgreSQL Backend

Quick Start

from pydantic_ai import Agent
from pydantic_ai_todo import create_todo_toolset

agent = Agent(
    "openai:gpt-4o",
    toolsets=[create_todo_toolset()],
)

result = await agent.run("Create a todo list for building a REST API")

Use Cases

Project Planning Agents

Let agents decompose complex features into trackable task hierarchies with dependencies and subtasks.

Development Workflow

Automate sprint planning — agents create, prioritize, and track tasks with dependency-aware ordering.

Task Decomposition

Break vague requests into concrete, actionable steps with clear completion criteria and blocking relationships.

Multi-User Apps

Use PostgreSQL backend for persistent, session-isolated task management across multiple users and conversations.

Ready to build your first production AI agent?

Open-source tools, battle-tested patterns, zero boilerplate. Configure your stack and ship in minutes — not months.

Build Your AI App