sisyphus

Orchestrate multi-agent workflows until all verification checklist items pass.

5|2|Updated Jan 15, 2026
One-click install
npx skills add https://github.com/sungmanch/single-person-company --skill sisyphus-sungmanch
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sisyphus
Source: https://github.com/sungmanch/single-person-company/tree/main/skills/sisyphus
Command: npx skills add https://github.com/sungmanch/single-person-company --skill sisyphus-sungmanch

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

This Skill addresses the challenge of ensuring complex, multi-step tasks are fully completed by providing a persistent, self-correcting orchestration system that delegates to specialized agents and verifies all conditions are met before stopping.

Core Features & Use Cases

  • Persistent Task Execution: Guarantees that tasks are not abandoned until all sub-tasks are finished.
  • Agent Delegation: Leverages a suite of specialized agents (e.g., oracle, librarian, frontend-engineer) for efficient task handling.
  • Parallel Processing: Optimizes workflow by running independent tasks concurrently.
  • Verification Checklist: Implements a rigorous checklist to ensure functionality, test coverage, error resolution, and quality before completion.
  • Background Task Management: Efficiently handles long-running operations like installations and builds.
  • Use Case: Managing a software development project where features must be implemented, tested, and documented, with the system continuing to work on any incomplete items until the entire project is delivered.

Quick Start

Use the sisyphus skill to orchestrate the development of a new user authentication module, ensuring all backend and frontend components are implemented, tested, and documented.

Frequently Asked Questions about sisyphus

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

FAQPage Schema
What is multi-agent workflow automation for persistent task execution?

Multi-agent workflow automation is a system that delegates complex tasks to specialized subagents and persists until all sub-tasks are finished. It uses parallel processing and background task management to ensure no work is abandoned before completion.

How do I ensure complex software development tasks are fully completed and verified?

To ensure tasks are fully completed, you can use an orchestration system that enforces a strict verification checklist. It checks for task completion, functionality, test coverage, error resolution, and quality before allowing the workflow to stop.

Can I run background and foreground tasks concurrently in a multi-agent orchestration system?

Yes, you can run background and foreground tasks concurrently. This multi-agent orchestration system optimizes workflows by managing long-running background operations like installations and builds while executing independent foreground tasks in parallel.

What is the best way to orchestrate end-to-end project execution across specialized agents?

The best way to orchestrate end-to-end project execution is to use a persistent system that delegates to specialized subagents like frontend-engineer or oracle. This approach manages the entire workflow from feature implementation to testing and documentation.

Does persistent task management work for continuous improvement cycles in software engineering?

Yes, persistent task management works for continuous improvement cycles. The system is designed to self-correct and continue working on incomplete items, enforcing verification for functionality and errors until the entire project is delivered.

When do I need a verification checklist for multi-agent task orchestration?

You need a verification checklist when managing complex, multi-step workflows that require guaranteed completion. It enforces strict conditions for functionality, tests, errors, and quality, preventing the orchestration system from stopping prematurely.