Propose Action
Action
Layer Inputs
Scenarios
Who has authority over this action? First = actor, rest = chain.
⚡ Mid-Commit World State Change
Simulate world state drifting to DRIFT between evaluate and commit — tests the exact execution boundary
Truth / Validity
Human State
World State
Authority
Pre-built scenarios — click to load into the action builder, then run governance.
Governance Flow
L0
Irreversibility Boundary
—
L1
Truth / Validity
—
L2
Action Classification
—
L3
Human State
—
L4
World State
—
L5
Objective Authority
—
L6
Policy / Constraint
—
L7
Execution Boundary (CLEAR)
—
—
T
Truth
—
A
Authority
—
H
Human
—
W
World
—
R
Risk
—
System Status & Audit
Health
Audit Log
FORGE
Evaluability Score
—
GROUND Status
—
Chain Integrity
—
Basis State
—
Admissible Path
—
System Uptime
—
📋
Run an action to see the audit trail
Run adversarial scenarios against CLASPION live. All 53 scenarios across 6 attack families.
ALL
53 scenarios
HUMAN
5 scenarios
AI-SIDE
7 scenarios
WORLD
5 scenarios
AUTHORITY
6 scenarios
META
5 scenarios