Assumption Extractor
Identifies and evaluates hidden dependencies in plans or arguments to surface critical risks and load-bearing premises.
Assumption Extractor is a specialized critical thinking tool for Claude Code that stress-tests strategies and arguments by identifying unstated premises. It systematically scans for four types of assumptions—factual, causal, value-based, and definitional—and ranks them by their 'load-bearing' weight. By flagging the most dangerous assumptions that lack verification, it helps developers and decision-makers conduct effective pre-mortems, avoid blind spots, and validate logic before committing significant resources to a project.
主な機能
01Structured 'Assumption Map' output for clear risk visualization
02Load-bearing classification to identify premises critical to project success
03Bilingual support for both English and Korean logic analysis
04Systematic four-category scanning (Factual, Causal, Value, and Definitional)
05Identification of the 'Most Dangerous Assumption' based on verification gaps
060 GitHub stars
ユースケース
01Conducting a pre-mortem on technical strategies or project roadmaps
02Identifying blind spots in complex architectural decisions before implementation
03Stress-testing the logic of a business plan or argument before stakeholder review