skills.rest
ssccio
ssccioCOMMUNITY
Explore all skills by this author →
GitHub

ultrawork

Execute independent tasks in parallel with tiered subagent routing.

Updated Feb 20, 2026
Search Tags:#parallel execution#agent orchestration#build automation#background jobs#model routing#task throughput#Claude Code skills
One-click install
npx skills add https://github.com/ssccio/cc-on-web-base --skill ultrawork-ssccio
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ultrawork
Source: https://github.com/ssccio/cc-on-web-base/tree/main/skills/ultrawork
Command: npx skills add https://github.com/ssccio/cc-on-web-base --skill ultrawork-ssccio

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ultrawork solves the bottleneck of slow, sequential task completion by enabling parallel execution of independent work so you can reach throughput targets faster.

Core Features & Use Cases

  • Parallel execution with multiple agents: Run independent tasks simultaneously for higher throughput.
  • Background operations: Offload long-running steps to background execution to keep workflows moving.
  • Smart model routing via tiered agents: Use the configured agent tiers and explicitly provided models to reduce token cost while preserving quality.
  • Role in larger workflows: Acts as a parallel execution component used by persistence/autonomy modes like ralph and autopilot rather than being the full end-to-end completion system.

Quick Start

Ask the AI to run a set of independent tasks in parallel using ultrawork, with model routing explicitly specified for each subagent type.

Frequently Asked Questions about ultrawork

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

FAQPage Schema
How do I run independent build and test tasks in parallel to improve throughput?▼

You can run independent build, test, and packaging tasks in parallel using tiered subagent routing to optimize token cost and execution responsiveness. This approach overcomes slow sequential workflows by executing high-volume operations concurrently.

What is the best way to offload long-running dependency installation to background execution?▼

Background execution allows you to offload long-running dependency installation and other high-volume operations, keeping workflows moving. This is achieved by applying background-vs-foreground execution rules alongside tiered subagent model routing.

Can I use model routing to reduce token costs when automating multi-step workflows?▼

Yes, tiered subagent routing with explicitly provided model parameters reduces token costs while preserving output quality. This allows you to optimize cost and responsiveness across concurrent operations.

Does parallel execution work for all stages of a multi-step workflow?▼

Parallel execution applies only to independent tasks that can be safely run concurrently, such as build, test, packaging, and dependency installation. It acts as a component for persistence and autonomy modes rather than handling the entire end-to-end workflow.

When should I not use parallel agent orchestration for task execution?▼

You should avoid parallel agent orchestration for tasks that are dependent on each other or require sequential completion. This approach is designed strictly for independent operations like build and packaging that can safely run concurrently.

Related Skills

Explore More Software Engineering Skills→
SoufianouassifSoufianouassif

delegation

Delegate tasks to synchronous or asynchronous subagents for automated execution.

Community
Advanced
sartoris-digitalsartoris-digital

executing-plans

Execute implementation plans with sequential, parallel, autopilot, and Ralph modes.

Official
Advanced
kwt00kwt00

ultrawork

Execute multiple independent tasks in parallel without state management.

Community
Advanced
sigridjinethsigridjineth
20

ultrawork

Coordinates parallel execution of independent tasks across multiple GPUs and machines.

Community
Intermediate