rune-problem-solver

Match the framework to what is unknown: use 5 Whys for unknown root causes, a weighted criteria matrix for multi-option decisions, and pre-mortem for high-stakes irreversible choices. First classify the problem's Cynefin domain to determine how much analysis depth is warranted.

Cynefin classification sorts problems into Clear, Complicated, Complex, Chaotic, or Confused domains to match analysis depth to problem complexity. Complicated problems get analytical frameworks, while Complex problems need probe-sense-respond experimentation instead of deep upfront analysis.

Bias detection scans each problem for patterns like confirmation bias, sunk cost fallacy, and anchoring, then applies debiasing strategies such as seeking disconfirming evidence or zero-based evaluation. It also steel-mans the least-favored option to prevent strawman dismissals.

Use a pre-mortem for high-stakes, irreversible decisions where blind spots are costly. Assume the decision has already failed 12 months later, work backward to list causes, then categorize by likelihood and severity and build mitigation plans for the highest risks.

Frameworks add overhead, so applying deep analysis to Clear-domain problems wastes effort, and purely technical problems like algorithm selection skip steps like PESTLE and ethics checks. Output is also capped at three ranked solutions and five evaluation criteria to avoid choice overload.