execute

Execute one sub-task from a refined issue and stop after completion.

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

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Execute EXACTLY ONE sub-task from a refined issue and stop after completing it.

Core Features & Use Cases

  • Single-task execution: Completes exactly one sub-task per invocation, preventing drift across tasks.
  • Backend flexibility: Works with Linear and Jira backends through progressive disclosure and unified context handling.
  • Safe, verifiable workflow: Runs verification steps (typecheck, tests, lint) and reports results to guide the next iteration.

Quick Start

Use the execute skill to process the next sub-task. Example: /execute MOB-177

Frequently Asked Questions about execute

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

FAQPage Schema
How do I automate subtask execution for Linear or Jira issues?

To automate subtask execution for Linear or Jira, you run a single-loop process that identifies the next ready subtask, implements the code, verifies it with tests and lint, then commits and pushes. This approach processes exactly one subtask per invocation.

What is the best way to process a single subtask from a refined Linear issue?

The best way to process a single subtask from a refined Linear issue is using a focused execution workflow that isolates one subtask, implements it, runs typecheck and lint verification, and updates the issue context before stopping.

Can I run automated subtask execution in a loop for Jira backends?

Yes, you can run automated subtask execution in a loop for Jira backends. The execution workflow is designed to be invoked repeatedly, with each loop iteration handling exactly one subtask, verifying it, and pushing the commit.

Does subtask automation run verification checks before committing code?

Yes, subtask automation runs verification checks before committing code. The workflow executes tests, typecheck, and lint steps to ensure code quality, reports the results to guide the next iteration, and then commits and pushes.

Why does my automated workflow only complete one subtask and then stop?

Automated workflows complete exactly one subtask and stop to prevent drift across multiple tasks. By focusing on a single subtask per invocation, the workflow ensures safe, verifiable implementation before moving to the next ready item in the loop.