deepthink

Automate structured multi-step reasoning for open-ended analytical questions.

1|Updated Jan 12, 2026
One-click install
npx skills add https://github.com/akiselev/qt-llm-template --skill deepthink-akiselev
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: deepthink
Source: https://github.com/akiselev/qt-llm-template/tree/main/.claude/skills/deepthink
Command: npx skills add https://github.com/akiselev/qt-llm-template --skill deepthink-akiselev

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

Structured reasoning for open-ended analytical questions, enabling robust multi-step analysis without manual scaffolding.

Core Features & Use Cases

  • Automated multi-step thinking across planning, sub-agent design, and iterative synthesis.
  • Sub-agent design, critique, and revision to advance complex analyses.
  • Deterministic workflow with quick mode for faster, scoped reasoning.

Quick Start

Run the think script to start the structured reasoning workflow.

Frequently Asked Questions about deepthink

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

FAQPage Schema
How do I automate multi-step reasoning for open-ended analytical questions?

You can automate multi-step reasoning by running a Python script that orchestrates planning, sub-agent design, iteration, and formatting. This structured workflow handles taxonomy design and conceptual analysis without manual scaffolding.

What is structured reasoning and when do I need it for taxonomy design?

Structured reasoning is a deterministic workflow that breaks down open-ended analyses into planned sub-agent tasks and iterative synthesis. You need it for taxonomy design or trade-off exploration when the answer structure is unknown.

How do I set up sub-agent design and iteration for conceptual analysis?

You set up sub-agent design and iteration by executing the included Python script, which automates the orchestration across planning, critique, revision, and formatting to advance complex conceptual analyses.

Does this structured reasoning workflow require any external dependencies?

No external dependencies are required. The structured reasoning workflow operates self-contained using a Python script to manage the multi-step planning and sub-agent orchestration.

Can I use a quick mode for faster scoped analytical reasoning?

Yes, you can use quick mode for faster, scoped reasoning. This mode limits the multi-step structured analysis to provide quicker outputs when full iterative synthesis is not required.

What is the best way to explore trade-offs when the answer structure is unknown?

The best way to explore trade-offs with an unknown answer structure is using automated sub-agent design and iterative synthesis. This approach systematically plans and revises the analysis to map out complex conceptual spaces.