Can I adjust the depth of the planning interview?

Yes, the skill uses an adaptive check pattern that allows users to request additional rounds of questioning or deeper dives into specific technical areas during the interactive phase.

What is the difference between Interactive and Auto mode?

Interactive mode conducts a 'Deep Interview' to clarify technical decisions with the user, while Auto mode makes decisions autonomously based on the provided session context.

How does the skill determine task granularity?

The planning protocol follows a rule-of-thumb targeting one deliverable per approximately 30 minutes of work to ensure tasks remain manageable, independent, and easily testable.

Where are the design documents saved?

Design documents are automatically generated and saved within your project's directory under the path 'docs/plans/' for easy versioning and reference.

Ultrawork Planning

Name: Ultrawork Planning
Author: mnthe

bymnthe

•

Collaboration Tools

Decomposes complex project goals into structured implementation plans and actionable tasks within the Ultrawork framework.

About

Ultrawork Planning is a sophisticated workflow skill for Claude Code designed to bridge the gap between high-level requirements and granular execution. It analyzes repository context to perform complexity assessments, conducts context-aware 'Deep Interviews' to clarify architectural decisions, and generates comprehensive design documents. By decomposing work into independent, testable 30-minute tasks, it ensures a systematic and predictable approach to software development, whether operating in a guided interactive mode or a fully autonomous execution mode.

Key Features

Automated complexity analysis to determine planning depth
Support for both interactive orchestrator and autonomous planner modes
Comprehensive design document generation including rationale and testing strategies
2 GitHub stars
Granular task decomposition into testable 30-minute work units
Context-aware 'Deep Interview' protocol for architectural decision-making

Use Cases

Breaking down high-level feature requests into manageable task lists for AI agents
Documenting technical design decisions and project scope for team alignment
Architecting multi-module refactors or system-wide rewrites

About

Key Features

Automated complexity analysis to determine planning depth
Support for both interactive orchestrator and autonomous planner modes
Comprehensive design document generation including rationale and testing strategies
2 GitHub stars
Granular task decomposition into testable 30-minute work units
Context-aware 'Deep Interview' protocol for architectural decision-making

Use Cases

Breaking down high-level feature requests into manageable task lists for AI agents
Documenting technical design decisions and project scope for team alignment
Architecting multi-module refactors or system-wide rewrites