causal-modeling
Skill2 starsChanged 6 months agoApache-2.0View on GitHub
Campaign for building causal models — identify variables, map mechanisms, collect evidence, analyze interventions, validate models. Produces causal graphs in the wiki vault.
In this repository70 other files it gives its agents.
- alias-resolutionSkill
- argument-mappingSkill
- argument-synthesisSkill
- argument-visualizationSkill
- axis-extractionSkill
- axis-validationSkill
- causal-chain-querySkill
- claim-decompositionSkill
- claim-extractionSkill
- claim-page-creationSkill
- combination-enumerationSkill
- combination-mappingSkill
- concept-decompositionSkill
- concept-extractionSkill
- concept-page-creationSkill
- confidence-scoringSkill
- confidence-updateSkill
- consistency-checkingSkill
- contradiction-flaggingSkill
- counterfactual-reasoningSkill
- dimensional-analysisSkill
- dimension-discoverySkill
- dimension-page-creationSkill
- domain-scopingSkill
- edge-batch-creationSkill
- evidence-attachmentSkill
- evidence-collectionSkill
- evidence-linking-argSkill
- evidence-linkingSkill
- evidence-weighingSkill
- feedback-loop-detectionSkill
- gap-detectionSkill
- gap-prioritizationSkill
- hierarchy-constructionSkill
- hierarchy-visualizationSkill
- intervention-analysisSkill
- intervention-page-creationSkill
- knowledge-compilationSkill
- loop-documentationSkill
- matrix-exportSkill
- matrix-generationSkill
- mechanism-edge-creationSkill
- mechanism-mappingSkill
- merge-candidatesSkill
- model-gap-detectionSkill
- model-validationSkill
- novelty-scoringSkill
- ontology-buildingSkill
- ontology-exportSkill
- ontology-refinementSkill
- question-generationSkill
- rebuttal-documentationSkill
- relation-typingSkill
- seed-concept-searchSkill
- source-gatheringSkill
- strength-assessmentSkill
- strength-scoringSkill
- synthesis-reportSkill
- taxonomy-validationSkill
- validation-reportSkill
- The first 60 shown.
The file
We split this skill into the rules it gives an agent
4 rules. About 0 are common, 1 is distinctive, and 3 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈1
Distinctive · estimate
This author's own choices. Worth a look.
3
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 49
Call context-init at campaign start
- Very specific: line 50
Call context-checkpoint after each strategy completes
- Very specific: line 51
Call knowledge-compilation after each strategy </HARD-GATE>
- Distinctive: line 56
Correlation is not causation: Every causal edge must have mechanistic justification, not just statistical association.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Causal Modeling
Build causal models for research domains. Identifies variables, maps causal mechanisms, collects supporting evidence, analyzes potential interventions, and validates the resulting causal graph.
Manifest
| Level | Count | Skills |
| Strategy | 5 | variable-identification, mechanism-mapping, evidence-collection, intervention-analysis, model-validation |
| Tactic | 3 | counterfactual-reasoning, evidence-weighing, feedback-loop-detection |
| SOP | 10 | variable-page-creation, mechanism-edge-creation, evidence-linking, contradiction-flagging, confidence-scoring, intervention-page-creation, loop-documentation, model-gap-detection, causal-chain-query, validation-report |
Budget Table
| Metric | Small | Medium | Large |
| Variables identified | 8 | 20 | 40 |
| Causal edges created | 15 | 40 | 80 |
| Evidence pages linked | 10 | 30 | 60 |
| Interventions analyzed | 2 | 5 | 10 |
| Feedback loops documented | 1 | 3 | 6 |
Strategy Sequence (Reference, Not Prescription)
- variable-identification — identify key variables in the causal system
- mechanism-mapping — map causal mechanisms between variables
- evidence-collection — gather evidence supporting/refuting causal claims
- intervention-analysis — analyze what happens when variables are manipulated
- model-validation — validate the causal model for consistency and completeness
MCP Tools Used
vault_search— find existing variables and mechanismsvault_add_edge— create causal edges (derived_from, supported_by, contradicts)vault_query_graph— trace causal chainsvault_graph_stats— assess model coveragevault_lint— validate structural integrity
Context-Management
<HARD-GATE>
- Call `context-init` at campaign start
- Call `context-checkpoint` after each strategy completes
- Call `knowledge-compilation` after each strategy
</HARD-GATE>Guiding Principles
- Correlation is not causation. Every causal edge must have mechanistic justification, not just statistical association.
- Confounders are everywhere. Actively search for confounding variables that could explain observed relationships.
- Interventions reveal truth. The strongest evidence for causation comes from intervention studies.
- Feedback loops are the norm. Most real systems have circular causation. Document loops explicitly.
- Confidence is calibrated. Strong mechanism + strong evidence = high confidence. Weak either = low confidence.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.