counterfactual-causal-analysis

SkillDev tools

Lets your agent test causal claims by building counterfactual scenarios and finding which factors truly matter.

Available today. Use it from your connected AI after setup.

Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.

Then ask your AI: use the counterfactual-causal-analysis skill

About this skill

Intervene on factors, construct counterfactual worlds, test necessity/sufficiency, and identify load-bearing causal factors.

What this skill tells your AI

The instructions your AI receives, as published by yogsoth-ai/de-anthropocentric-research-engine in skills/counterfactual-causal-analysis/SKILL.md and read by ahel’s review.

Purpose

Intervene on factors, construct counterfactual worlds, test necessity and sufficiency, and identify load-bearing causal factors.

Input contract

required: [causal_claim, variables, intervention_target]
optional: [mechanism_graph, baseline_world, comparison_conditions]
constraints: [intervention and outcome must be measurable or explicitly qualitative]

Execution protocol

Do not perform called SOP operations inline; each loaded SOP owns its contract and thresholds.

  1. You MUST load skill extract-causal-structure to extract the causal structure. You MUST load skill identify-variables to type its variables.
  2. You MUST load skill construct-counterfactual to construct the counterfactual. You MUST load skill identify-load-bearing-factors to identify load-bearing factors.
  3. You MUST load skill search-minimal-flip to search the minimal conclusion flip. You MUST load skill evaluate-necessity-sufficiency to test necessity and sufficiency. If the conclusion should instead be tested by contradiction and counterexample, consider reductio-counterexample-analysis as the next tactic. Deviation: use qualitative counterfactuals when intervention data are unavailable, but mark the inference uncertainty.

Output contract

produces: [causal_structure, counterfactual_map, load_bearing_factors, minimal_flip, necessity_sufficiency_report]
delta_fields: [findings, evidence_updates, hypothesis_updates, uncertainties, decisions]

Thresholds and quality gates

  • Every intervention must state changed variables, held-fixed conditions, predicted outcome, and evidence basis.
  • Necessity and sufficiency are separate judgments; do not collapse them.

Failure and counterexamples

Do not call a factor necessary when an untested substitute can produce the outcome. Mark counterfactuals underdetermined when held-fixed conditions are unspecified.

Provenance map

  • resolved: counterfactual-probing
  • resolved: structural-counterfactual
  • resolved: necessity-sufficiency
  • resolved: systematic-factor-ablation
  • resolved: causal-necessity-testing
  • resolved: minimal-change-search

Preserved source criteria ledger

sourcesource linekindsource criterion
v4 architecturenode desctextualIntervene, construct counterfactual worlds, test necessity/sufficiency, identify load-bearing factors.

Context checkpoint / Delta notes

Append intervention worlds, held-fixed conditions, factor ranking, flip search, and necessity/sufficiency results.

Signals

GitHub stars
501
Forks
41
Last commit
Sep 2026
Advanced
Catalog kind
skill
Key
counterfactual-causal-analysis
Source
github.com/yogsoth-ai/de-anthropocentric-research-engine