refine

Break down issues into structured subtasks with local JSON outputs across Linear, Jira, and local backends.

1|Updated Jan 25, 2026
One-click install
npx skills add https://github.com/Tubular-Health/mobius --skill refine
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: refine
Source: https://github.com/Tubular-Health/mobius/tree/main/.claude/skills/refine
Command: npx skills add https://github.com/Tubular-Health/mobius --skill refine

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps teams translate vague issues into precise, executable subtasks with clear dependencies, enabling parallel work and predictable delivery.

Core Features & Use Cases

  • Automated sub-task generation: creates local JSON task files with a focused scope and a clear blocking order.
  • Backend-agnostic support: works with Linear, Jira, and Local backends; writes tasks locally for offline planning.
  • Guided workflow: covers issue definition, refinement, task aggregation, and verification.

Quick Start

Use refine MOB-123 to start breaking down an issue into subtasks.

Frequently Asked Questions about refine

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

FAQPage Schema
How do I break down a Linear issue into structured subtasks?

To break down a Jira issue into subtasks, the system fetches the parent issue and writes structured subtasks locally as JSON files. It validates the issue ID and detects the backend automatically to prepare tasks for offline planning.

Can I generate subtasks for offline project management planning?

Yes, you can generate subtasks for offline project management by writing structured JSON task files locally. The guided workflow operates across Local, Linear, and Jira backends to ensure executable tasks with clear dependencies are available without constant cloud syncing.

Does automated task breakdown work with both Jira and local backends?

Automated task breakdown works with Jira, Linear, and local backends. The system detects the backend automatically based on the provided issue ID, enabling backend-agnostic subtask generation that writes structured tasks as local JSON files for offline planning.

What is the best way to turn vague issues into executable subtasks with dependencies?

The best way to turn vague issues into executable subtasks is using a guided four-phase workflow that defines, refines, aggregates, and verifies tasks. This process translates broad issues into local JSON files featuring focused scopes and a clear blocking order for parallel work.

How do I validate issue IDs before generating local JSON subtasks?

Issue IDs are validated automatically as the first step of the task breakdown workflow. Before generating local JSON subtasks, the system checks the ID format and detects whether it belongs to a Linear, Jira, or local backend to ensure accurate parent issue fetching.