sisyphus

Orchestrate feature development from requirements through implementation to quality gate.

6|Updated Jan 16, 2026
One-click install
npx skills add https://github.com/kok1eee/o-m-cc --skill sisyphus-kok1eee
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: sisyphus
Source: https://github.com/kok1eee/o-m-cc/tree/main/skills/sisyphus
Command: npx skills add https://github.com/kok1eee/o-m-cc --skill sisyphus-kok1eee

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill automates and enforces a specification-driven development workflow that prevents premature completion claims and ensures features move from discovery through design, implementation, verification, and a quality gate without human micromanagement.

Core Features & Use Cases

  • Multi-phase orchestration: Chains independent skills for discovery, design, task decomposition, implementation, verification, and quality gating.
  • Agent Teams & Verification: Uses peer agents and an adversarial Verifier plus a verification skill for two-stage evidence-based checks, reducing author bias.
  • Headless & Monitor support: Supports fully automated (headless) runs, long-running test streaming via Monitor, and task tracking with TaskCreate/TaskUpdate lifecycle.
  • Use Case: Ideal for implementing new features or complex refactors where requirements must be validated, tasks decomposed, and each change independently verified before merge.

Quick Start

Start a full Sisyphus run by invoking the sisyphus skill with a concise feature description, for example: /o-m-cc:sisyphus "Implement authentication flow".

Frequently Asked Questions about sisyphus

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

FAQPage Schema
How do I automate end-to-end feature development from requirements to verification?

End-to-end feature development can be automated using a specification-driven workflow that orchestrates discovery, design, task decomposition, implementation, and quality gating without human micromanagement. This prevents premature completion claims by enforcing structured verification loops.

What is adversarial verification in software engineering workflows?

Adversarial verification is a two-stage evidence-based checking process utilizing peer agents and a dedicated Verifier to independently validate implementation against requirements, significantly reducing author bias before passing a quality gate.

How do I decompose complex refactoring tasks for parallel agent execution?

Complex refactoring tasks can be decomposed through a multi-phase orchestration workflow that breaks down specifications into independent tasks, enabling parallel agent execution and continuous task tracking via lifecycle updates.

Can I run automated feature implementation and testing fully headless?

Automated feature implementation supports fully headless runs for continuous execution, while also offering monitor-based long-running test streaming to observe real-time progress and verification loops interactively.

When do I need a quality gate for new feature implementation?

A quality gate is needed when implementing new features or complex fixes where requirements must be validated and changes independently verified through an adversarial process, ensuring no premature completion claims occur before merge.