hypothesis

Hypothesis generation for root-cause analysis involves creating 2-4 candidate hypotheses with explicit evidence. You then perform adversarial falsification and environment-fit checks to identify the strongest surviving cause. This structured method ensures your investigation yields actionable targets.

A Tree of Thoughts for causation analysis is a structured evaluation method that branches out potential root causes. It allows you to systematically test and falsify various paths against supporting evidence, ensuring the surviving hypothesis accurately reflects the investigation environment.

To test investigation evidence against environment-fit constraints, you apply your Phase 2-3 data to validate root causes. This involves checking candidate hypotheses against environment documentation to ensure the identified causation aligns with system precedents and actual configurations.

The best way to plan downstream targets after root-cause identification is to classify the issue type and establish actionable objectives. By surviving the best hypothesis from your falsification tests, you directly guide Phase 5 planning and gate decisions for the investigation.

Yes, you can use inv_context.json data for Phase 4 hypothesis generation. The Skill specifically loads investigation state and extracts Phase 2-3 data from this file to build falsifiable hypothesis sets and perform environment-fit assessments.

Root-cause analysis often fails to produce actionable targets when hypotheses are not falsifiable. Without applying a Tree of Thoughts evaluation and adversarial falsification, you lack the evidence-based validation needed to confidently classify issues and plan next steps.