think

Audits conversations for assumptions, biases, unchecked facts, and premature convergence.

Updated May 3, 2026
One-click install
npx skills add https://github.com/spikelab/multiplai-cc-mktplace --skill think-spikelab
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: think
Source: https://github.com/spikelab/multiplai-cc-mktplace/tree/main/plugins/multiplai-dev/skills/think
Command: npx skills add https://github.com/spikelab/multiplai-cc-mktplace --skill think-spikelab

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Conversations and decisions often lock onto a single framing without testing assumptions, checking facts, or considering alternatives. This Skill stress-tests the thinking already on the table so blind spots, biases, and weak reasoning surface before they become costly decisions. ## Core Features & Use Cases - Quick Audit Mode: Reads the current conversation and delivers a structured report covering assumptions (with certainty/impact ratings), detected cognitive biases, unchecked facts, alternative framings, and convergence alerts. - Focused Mode: Runs an interactive 10-15 exchange dialogue on a specific problem, applying Skeptic, Sage, and Strategist lenses one question at a time. - Deep Mode: Guides high-stakes or irreversible decisions through a seven-phase structured process with a persistent tracking document, including root cause analysis, divergent ideation (minimum 10 options), premortem stress testing, and a quality gate. - Use Case: Before committing to a career change, run a deep session to generate framings, document success criteria, produce 10+ options, and record kill conditions for the chosen path. ## Quick Start Ask the assistant to run /think with no arguments to audit the current conversation, or say "check my thinking on this decision" to get a structured critical review.

Frequently Asked Questions about think

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

FAQPage Schema
How do I audit my conversation for assumptions and biases?

Invoke the skill with no arguments to run Quick mode, which reads the entire conversation and returns a structured audit. The report rates each assumption by certainty and impact, names detected biases with test questions, and lists unchecked facts worth verifying.

What cognitive biases does a thinking audit check for?

The bias checklist covers inertia biases like sunk cost and status quo, framing biases like anchoring and base rate neglect, confidence biases like overconfidence and planning fallacy, plus confirmation bias, survivorship bias, groupthink, and action biases.

How do I structure a high-stakes decision process?

Use Deep mode, which walks through seven phases: understanding, root cause analysis, divergent ideation with at least 10 options, stress testing via premortem and scenarios, evaluation against ranked success criteria, a quality gate, and a conclusion with kill conditions.

Can I use critical thinking analysis on one specific problem?

Yes, Focused mode handles a bounded problem through 10-15 interactive exchanges. It applies the Skeptic, Sage, and Strategist lenses one question at a time, then summarizes what was clarified, which assumptions were tested, and what options emerged.

When should I not use a quick thinking audit?

Quick mode is non-interactive and only examines information already in the conversation, so it cannot surface information you have not shared. For irreversible or complex decisions, use Deep mode; for extracting hidden information first, pair it with an interview-style skill.