---
name: yogsoth-ai/feedback-loop-detection
source: https://app.decimal.ai/s/yogsoth-ai-feedback-loop-detection@1/SKILL.md
source_sha256: 7b5cf523d8c9
---

# 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

<HARD-GATE>
≥1 loop identified and classified (or explicit confirmation that no loops exist in the current subgraph) per invocation.
</HARD-GATE>

<!-- BEGIN available-tables (generated) -->

## Available SOPs

Optional, no fixed order; the final leaf is always a sop.

| SOP | When to use |
| --- | --- |
| causal-chain-query | SOP for tracing causal chains — follow edges from cause to effect through intermediate variables. |
| loop-documentation | SOP for documenting a feedback loop — classify, describe dynamics, identify break points. |
| mechanism-edge-creation | SOP for creating a causal mechanism edge — documents how one variable causes changes in another. |

<!-- END available-tables (generated) -->