juggle

Automates task management for AI agent loops via CLI commands.

15|1|Updated Nov 12, 2025
One-click install
npx skills add https://github.com/ohare93/juggle --skill juggle
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: juggle
Source: https://github.com/ohare93/juggle/tree/main/skills/juggle
Command: npx skills add https://github.com/ohare93/juggle --skill juggle

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Juggle simplifies coordinating AI agent loops by providing a structured CLI workflow for creating, updating, and tracking task units (balls) and groupings (sessions), with clear acceptance criteria and an auditable progress log.

Core Features & Use Cases

  • CLI-first orchestration: list, show, plan, update, progress, and sessions commands to manage balls and sessions without editing files directly.
  • Entities: Balls as tasks with intent, acceptance criteria, state, tags, and optional model size; Sessions as context and progress journals.
  • Non-interactive / headless mode support for automation, robust dependency handling, and predictable outcomes.

Quick Start

Use the juggle skill to manage tasks.

Frequently Asked Questions about juggle

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I manage AI agent loops with a CLI workflow?

You can manage AI agent loops by using a CLI-first workflow to create, update, and track tasks. This approach organizes projects into a dedicated directory, grouping related work into sessions and logging progress without editing files directly.

What is the best way to track task dependencies in non-interactive agent runs?

Tracking task dependencies in non-interactive agent runs is handled through structured CLI commands that manage task states and acceptance criteria. This enables robust dependency handling and predictable outcomes for headless automation.

How do I organize AI project tasks with acceptance criteria and session journals?

Organizing AI project tasks involves creating task units called balls that contain intent, acceptance criteria, state, and tags. Related work is grouped into sessions that act as context and progress journals for auditable tracking.

Can I automate task tracking for headless AI agents without editing files directly?

Yes, you can automate task tracking for headless AI agents using a non-interactive mode. A CLI provides list, show, plan, update, and sessions commands to manage tasks and dependencies predictably without direct file editing.

Does juggle work with projects that need structured task management and progress logging?

Juggle works with projects requiring structured task management by applying a dedicated directory structure. It tracks task units with acceptance criteria and groups related work into sessions, enabling planning and auditable progress logging.

Why use a CLI to orchestrate tasks instead of managing files directly for AI agents?

Using a CLI to orchestrate tasks provides a structured workflow that avoids direct file editing, reducing errors. It standardizes task creation, dependency management, and progress logging, ensuring predictable outcomes for AI agent loops.