ultra-think

Analyzes complex problems through multi-perspective evaluation with calibrated confidence levels.

1|Updated Mar 8, 2026
One-click install
npx skills add https://github.com/kubrickcode/cine-mirror --skill ultra-think-kubrickcode
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: ultra-think
Source: https://github.com/kubrickcode/cine-mirror/tree/main/.agents/skills/ultra-think
Command: npx skills add https://github.com/kubrickcode/cine-mirror --skill ultra-think-kubrickcode

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Complex architectural decisions and strategic planning often suffer from shallow analysis, overconfident recommendations, and hidden assumptions. This Skill enforces rigorous multi-dimensional thinking that surfaces root causes, generates genuinely distinct approaches, and honestly weighs trade-offs before recommending a path. ## Core Features & Use Cases - Structured Deep Analysis: Breaks problems into core challenges, hidden constraints, and at least three genuinely distinct approaches with weighted trade-offs. - Calibrated Confidence: Defaults to Medium confidence, forbids overconfident language, and requires explicit uncertainty statements distinguishing confirmed facts from inference. - Self-Verification Checklist: Validates that every approach has real weaknesses, uncertainties are substantive, and recommendations acknowledge what is being given up. - Use Case: When deciding between microservices and a monolith for a growing application, invoke this Skill to receive a structured comparison of architectural approaches with honest trade-off analysis and a justified recommendation. ## Quick Start Use ultra-think to analyze whether we should migrate our monolithic application to microservices.

Frequently Asked Questions about ultra-think

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

FAQPage Schema
How do I get a deeper analysis of a complex technical decision?▼

Invoke the Skill with your decision problem as the argument. It produces a structured analysis covering the core challenge, hidden constraints, at least three distinct approaches with trade-offs, and a recommendation with stated confidence and uncertainties.

What kinds of problems is deep multi-perspective analysis best for?▼

It fits complex architectural decisions, strategic planning, and problems where multiple viable approaches exist and trade-offs are unclear. Simple problems receive proportionally concise analysis rather than forced depth.

How does the analysis avoid overconfident recommendations?▼

Confidence defaults to Medium and only rises when one approach is Pareto-dominant or a hard constraint eliminates alternatives. Forbidden language like "obviously" or "no-brainer" is banned, and claims are marked as confirmed, likely, or uncertain.

When should I not use structured deep analysis?▼

Avoid it for simple, well-understood tasks with one obvious solution, where the overhead of multi-approach evaluation adds no value. It is designed for genuinely ambiguous decisions, not routine implementation work.

What output format does the analysis produce?▼

The output follows a fixed Markdown structure: Problem Analysis, Approaches with pros/cons/risks and confidence, a Recommendation with rationale and reversal conditions, and an Uncertainties section listing what remains unknown.