Installs into .claude/skills of the current project.
Are you the author of Key Assumptions Check?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/yogsoth-ai-key-assumptions-check)
---
name: key-assumptions-check
description: "Military ACT: systematically enumerate all assumptions, classify by type, and evaluate evidence strength supporting each."
execution: subagent
prompt: ./prompt.md
input: artifact (string), artifact_type (string)
used-by: [red-teaming]
---
# Key Assumptions Check
CIA/military structured analytic technique for surfacing and evaluating assumptions.
## Execution
Subagent — spawned via subagent-spawning/spawn-agent.
## Why Subagent
Assumption enumeration requires neutral analytical stance. The agent must surface assumptions without defending or attacking them.
## Input
- **artifact**: The complete artifact to analyze
- **artifact_type**: Type of artifact (hypothesis, claim, idea, etc.)
## Output
- **assumptions**: Complete list with classification (explicit/implicit, foundational/operational)
- **evidence_strength**: Rating for each assumption (strong/moderate/weak/unsupported)
- **criticality**: How important each assumption is to the overall argument
