resolving-uncertainty

Resolving uncertainty converts unknowns into an Uncertainty Register—a prioritized list scored by Impact, Evidence, Urgency, and Effort. Each uncertainty becomes a testable hypothesis with observation tasks, letting you decide what to investigate first and what evidence you need to move forward.

Assumptions are beliefs treated as facts; risks are known threats with probability; unknowns are gaps in knowledge. This skill decomposes all three into single questions, scores them by business and effort factors, and produces an observation backlog so you investigate systematically instead of reactively.

The skill scores each uncertainty across Impact, Evidence, Urgency, and Effort, then ranks them. High-impact, low-effort items surface first, letting you resolve blockers quickly and gather evidence before committing resources to major decisions like MVP scope or architecture.

Yes. The skill produces an Uncertainty Register, Prioritized Top-N list, Observation Backlog, and optional uncertainty_plan.json. Export formats support decision documentation, team handoff, and verification via included scripts so investigations stay traceable.

When outcomes depend on unknown variables—feature adoption, performance thresholds, user behavior—hypothesis testing with observation tasks reduces risk. This skill converts vague assumptions into testable questions with clear success criteria and evidence collection methods.

Yes. The skill includes Python scripts for decomposing uncertainties, scoring, and exporting results. It integrates into Python workflows and supports verification of uncertainty registers through provided automation.