ultrawork

Parallelize independent agent tasks across model tiers with background execution.

Updated Apr 10, 2026
One-click install
npx skills add https://github.com/lpjhelder/kronon-forge --skill ultrawork-lpjhelder
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: ultrawork
Source: https://github.com/lpjhelder/kronon-forge/tree/main/skills/ultrawork
Command: npx skills add https://github.com/lpjhelder/kronon-forge --skill ultrawork-lpjhelder

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Ultrawork eliminates wasted time from serial task execution by running multiple independent agents concurrently and routing each task to the appropriate model tier, enabling high-throughput completion without adding persistence or heavy verification overhead.

Core Features & Use Cases

  • Parallel dispatch: Launch many independent agent tasks simultaneously to reduce wall-clock time.
  • Model tier routing: Route tasks to LOW, STANDARD, or THOROUGH tiers to balance cost and capability.
  • Background execution: Use run_in_background for long-running operations like installs, builds, and test suites.
  • Lightweight verification: Perform minimal checks when invoked directly (build/typecheck, affected tests, no new errors).
  • Use case: Run a trio of independent tasks—implement an endpoint, add integration tests, and perform a build—at once to shorten delivery time.

Quick Start

Use the ultrawork skill to execute multiple independent tasks in parallel by delegating each to the correct tier and marking long operations to run in the background.

Frequently Asked Questions about ultrawork

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

FAQPage Schema
How do I run parallel agent tasks to reduce execution time?

Parallel agent tasks reduce execution time by dispatching independent operations concurrently. You can launch unrelated tasks simultaneously to shorten wall-clock time, avoiding the delays of serial execution.

Can I route tasks to different model tiers for parallel execution?

Model tier routing allows parallel execution to assign tasks to LOW, STANDARD, or THOROUGH tiers. This balances cost and capability by matching task complexity to the appropriate model tier.

How do I execute long-running builds and test suites in the background?

Long-running builds and test suites execute in the background using run_in_background. This delegates lengthy operations like installs and test suites so they process without blocking other concurrent tasks.

What is lightweight verification for concurrent agent workflows?

Lightweight verification for concurrent agent workflows performs minimal checks when invoked directly. It validates builds, typechecks, and affected tests to ensure no new errors without heavy overhead.

When should I parallelize independent tasks instead of running them serially?

Parallelize independent tasks instead of running them serially when multiple operations lack strict dependency ordering. This applies to small fixes, feature implementations, builds, and tests to achieve high-throughput completion.

Do I need explicit model parameters to delegate tasks for concurrent execution?

Concurrent task delegation requires explicit model parameters to route operations correctly. You must specify the model tier for each task to balance cost and capability during parallel dispatch.