ultrawork

Orchestrate parallel execution of independent tasks with lightweight evidence and acceptance criteria.

Updated Apr 23, 2026
One-click install
npx skills add https://github.com/wmasfoe/notch-pet --skill ultrawork-wmasfoe
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ultrawork
Source: https://github.com/wmasfoe/notch-pet/tree/main/.codex/skills/ultrawork
Command: npx skills add https://github.com/wmasfoe/notch-pet --skill ultrawork-wmasfoe

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ultrawork enables parallel execution of independent tasks with lightweight evidence and decision discipline to accelerate high-throughput task completion.

Core Features & Use Cases

  • Parallel task execution with context discipline for high-throughput workloads
  • Lightweight evidence lanes and upfront acceptance criteria to guide execution
  • Flexible self-vs-delegate workflow management for fast, bounded tasks
  • Use Case: when you have multiple independent tasks that can run in parallel without heavy persistence guarantees

Quick Start

Provide a task intent and required constraints, then run ultrawork to start parallel execution.

Frequently Asked Questions about ultrawork

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

FAQPage Schema
What is parallel task execution with lightweight evidence?

Parallel task execution with lightweight evidence runs independent tasks concurrently while applying upfront acceptance criteria and decision discipline to accelerate high-throughput completion. It provides verification lanes without delaying core delivery or requiring heavy persistence guarantees.

How do I start parallel task execution for independent engineering workflows?

To start parallel task execution, provide a task intent and required constraints, then run the workflow to initiate concurrent lanes. The system applies self-vs-delegate decision making and defined pass/fail criteria to manage the parallel workload.

Can I use parallel execution lanes for tasks without full persistence guarantees?

Yes, parallel execution lanes are designed for independent tasks needing faster results without full persistence guarantees. It uses lightweight evidence and upfront acceptance criteria to verify bounded tasks without delaying core delivery.

What's the best way to manage self-vs-delegate decision making for concurrent tasks?

The best way to manage self-vs-delegate decisions is defining upfront task intent and pass/fail criteria before running parallel lanes. This flexible workflow management ensures fast, bounded tasks complete with lightweight verification.

When should I not use parallel task execution for engineering workloads?

You should not use parallel task execution for dependent tasks requiring heavy persistence guarantees or when explicit acceptance criteria and parallel lanes offer no advantage. It is built for independent, concurrent workloads needing faster results.