skills.rest
bahayonghang
bahayonghangCOMMUNITY
Explore all skills by this author →
GitHub

sisyphus

Orchestrate multi-step development tasks with TODO tracking and agent delegation.

16|1|Updated Jan 2, 2026
Search Tags:#delegation#phases#parallel-execution#orchestrate#task-planning#TODO-tracking
One-click install
npx skills add https://github.com/bahayonghang/my-claude-code-settings --skill sisyphus-bahayonghang
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: sisyphus
Source: https://github.com/bahayonghang/my-claude-code-settings/tree/main/skills/sisyphus
Command: npx skills add https://github.com/bahayonghang/my-claude-code-settings --skill sisyphus-bahayonghang

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

Sisyphus coordinates and delegates complex development tasks, enabling aggressive parallel execution and coordination across multiple subtasks. It solves the overhead of planning, delegation, and multi-file coordination in large projects.

Core Features & Use Cases

  • Aggressive parallelization: run independent subtasks concurrently to reduce lead time.
  • Delegation-first: hand off specialized work to appropriate agents rather than attempting to do everything.
  • TODO-driven: maintain all work via a centralized TODO list for traceability.
  • Phase-based workflow: supports intent judgment, codebase assessment, exploration, and implementation phases.
  • Multi-file coordination: coordinates changes across several files and modules.

Use cases include large feature development, significant refactors, research+implementation tasks, and cross-team coordination.

Quick Start

Start Sisyphus by describing your multi-step task. The system will generate a TODO-driven plan and delegate subtasks to specialized agents like @explore, @librarian, and @oracle. Monitor and update the TODO list as tasks complete.

Frequently Asked Questions about sisyphus

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

FAQPage Schema
How do I orchestrate parallel execution for complex multi-step development tasks?▼

Parallel execution for complex development tasks is orchestrated by generating a centralized TODO-driven plan and delegating subtasks to specialized agents. This structured approach enables concurrent processing of independent subtasks to reduce lead time.

What is the best way to coordinate multi-file refactoring across multiple modules?▼

Multi-file coordination for refactoring is best handled through a delegation-first workflow that assesses the codebase and hands off specialized work to appropriate agents. A centralized TODO list maintains traceability across all changed files and modules.

Can I use a structured phase workflow for large feature development and research tasks?▼

Yes, large feature development and research tasks use a structured three-phase workflow supporting intent judgment, codebase assessment, exploration, and implementation. This ensures safe delivery by coordinating changes across multiple files and teams.

How do I manage a centralized TODO list when delegating specialized coding work to agents?▼

A centralized TODO list tracks delegated specialized coding work by maintaining all task states in one place. You monitor and update the TODO list as agents like @explore, @librarian, and @oracle complete their assigned subtasks.

When should I use delegation and parallel task planning for software engineering projects?▼

Delegation and parallel task planning should be used for significant refactors, cross-team coordination, and large feature development. It solves the overhead of planning and multi-file coordination by running independent subtasks concurrently.

Related Skills

Explore More Software Engineering Skills→
phoenixvcphoenixvc

orchestrate

Manage multi-step AI agent workflows across a five-phase lifecycle.

Official
Advanced
porcupine-mdporcupine-md
11

orchestrating-feature

Coordinate multi-phase feature delivery across specialized AI agents.

Official
Advanced
jivebreaddevjivebreaddev

sisyphus

Coordinates specialized AI agents for multi-step tasks with argus verification.

Community
Advanced
joaquimscostajoaquimscosta
21

workflow-orchestration

Coordinate multi-agent parallel execution for complex multi-step tasks.

Community
Advanced