Feedback Loop Detection

Tactic for identifying circular causation — detect feedback loops, classify as reinforcing or balancing, document loop structure.

yogsoth-ai Updated 512 repo stars

File contents

Feedback Loop Detection

Identify circular causation in the causal graph. Most real systems have feedback loops — they must be documented explicitly.

Available SOPs

  • causal-chain-query — trace paths looking for cycles
  • loop-documentation — document identified loops
  • mechanism-edge-creation — create edges that close loops

Guiding Principles

  • Loops are normal. Don't assume acyclicity. Real causal systems almost always have feedback.
  • Classify loops. Reinforcing (positive feedback, amplification) vs balancing (negative feedback, homeostasis).
  • Time delays matter. A loop with a 10-year delay behaves differently from one with a 10ms delay.
  • Dominant loops. In systems with multiple loops, identify which loop dominates behavior.
  • Break points. Identify where interventions could break harmful loops.

Minimum Yield

yogsoth-ai/knowledge-structuring commit d74abbcf6d

Frequently asked questions

npx get-skill add yogsoth-ai/feedback-loop-detection-2