Will Ultrawork make decisions without asking me?

Yes, to maintain high intensity and momentum, the skill shifts Claude's behavior to make reasonable engineering decisions independently rather than stopping to ask for clarification on minor points.

What is Sisyphus Mode in the Ultrawork skill?

Sisyphus Mode is a high-intensity operational state where Claude Code is instructed to work persistently, rolling the 'boulder' of tasks until the entire todo list is completed without premature stopping.

How does the Todo-Continuation Enforcer work?

The enforcer is a hook that monitors the agent's status; if the agent attempts to stop while todos are still pending, it is automatically prompted to continue working on the remaining items.

How does this skill ensure code quality?

Despite the high intensity, quality is maintained through mandatory auto-verification and the execution of LSP diagnostics after every single file change.

How do I activate Ultrawork mode?

You can trigger the mode by using the commands 'ulw', 'ultrawork', or 'uw' within your Claude Code environment.

Ultrawork (Sisyphus Mode)

Name: Ultrawork (Sisyphus Mode)
Author: zephyrpersonal

byzephyrpersonal

0•

Productivity & Workflow

Activates a high-intensity task execution mode for Claude Code that enforces persistent progress and exhaustive verification until all objectives are met.

Ultrawork (also known as Sisyphus Mode) is a specialized productivity skill designed for complex, multi-step engineering tasks that require obsessive attention to detail and unwavering momentum. It transforms Claude's behavior from a standard assistant into a proactive agent that minimizes back-and-forth questioning, utilizes maximum search depth for research, and implements a 'todo-continuation-enforcer' to ensure no task is left unfinished. By automating diagnostics and verification after every change, it ensures high-quality output while maintaining an aggressive pace of development, making it ideal for deep refactoring and autonomous feature implementation.

Key Features

01Maximum Search Depth: Automatically applies 'very thorough' parameters to all exploration and documentation searches.

02Continuous Auto-Diagnostics: Runs LSP diagnostics and verification checks immediately after every file modification.

030 GitHub stars

04Todo-Continuation Enforcement: Prevents the agent from stopping until every item on the task list is marked as complete.

05Autonomous Decision Making: Reduces friction by making reasonable engineering decisions instead of pausing for clarification.

06Atomic Task Decomposition: Instantly breaks down complex user requests into granular, actionable todo lists.

Use Cases

01Refactoring large legacy codebases where exhaustive verification and persistent execution are required.

02Deep-dive debugging sessions that involve navigating complex documentation and multi-file dependencies.

03Autonomous feature development where the user defines a goal and needs the agent to work through the entire implementation cycle.

Key Features

01Maximum Search Depth: Automatically applies 'very thorough' parameters to all exploration and documentation searches.

02Continuous Auto-Diagnostics: Runs LSP diagnostics and verification checks immediately after every file modification.

030 GitHub stars

04Todo-Continuation Enforcement: Prevents the agent from stopping until every item on the task list is marked as complete.

05Autonomous Decision Making: Reduces friction by making reasonable engineering decisions instead of pausing for clarification.

06Atomic Task Decomposition: Instantly breaks down complex user requests into granular, actionable todo lists.

Use Cases

01Refactoring large legacy codebases where exhaustive verification and persistent execution are required.

02Deep-dive debugging sessions that involve navigating complex documentation and multi-file dependencies.

03Autonomous feature development where the user defines a goal and needs the agent to work through the entire implementation cycle.