counterfactual-scenario-construction

$npx mdskill add yogsoth-ai/de-anthropocentric-research-engine/counterfactual-scenario-construction

Constructs counterfactual scenarios by altering specified factors and analyzing outcomes

  • Solves tasks requiring analysis of alternative realities with modified inputs
  • Relies on subagent execution and internal reasoning for consistency
  • Evaluates cascading effects and maintains logical coherence in scenarios
  • Returns structured results including conclusions and consistency status
SKILL.md
.github/skills/counterfactual-scenario-constructionView on GitHub ↗
---
name: counterfactual-scenario-construction
description: Construct precise, internally consistent counterfactual scenarios where specified factors are altered, then reason about the resulting conclusion.
execution: subagent
prompt: ./prompt.md
input: artifact (string), factor_to_change (string), change_specification (string)
used-by: [counterfactual-probing]
---

# Counterfactual Scenario Construction

Builds precise counterfactual worlds where specified factors differ from actuality.

## Execution

Subagent — spawned via subagent-spawning/spawn-agent.

## Why Subagent

Scenario construction requires careful reasoning about cascading effects of changes while maintaining internal consistency.

## Input

- **artifact**: The original artifact
- **factor_to_change**: Which factor to alter
- **change_specification**: How to alter it (remove, weaken, strengthen, invert)

## Output

- **scenario**: The counterfactual world description
- **conclusion_status**: holds/weakened/flipped/indeterminate
- **cascading_effects**: What else changes as a result
- **consistency_check**: Whether the scenario is internally consistent

## Budget

One unit = one scenario construction per factor change.
More from yogsoth-ai/de-anthropocentric-research-engine