Pattern Proposal: State-Machine-Constrained Action Engine

Summary

Restricts agent-driven governance actions to strictly valid transitions registered in a deterministic state registry.

Candidate Pattern

• ID: state-machine-constrained-action-engine
• Classification: governance
• Maturity: established
• Type: implemented

Observed Signals

No specific signals listed.

Evidence

Supporting Evidence

• src/cas_runner/governance/governance_action_engine.py (TransitionRegistry, GovernanceActionEngine) - Registry enforcing valid state transitions for governance records
• src/cas_runner/governance/models.py (GovernanceAction, GovernanceException) - Typed models defining discrete governance lifecycle mutations

Scope Assessment

• Evidence Strength: high
• Confidence: high
• Applies to: ideas-to-life-cas

Rationale

• Intent: Prevent LLM or workflow agents from transitioning architectural assets into invalid lifecycle states (e.g. promoting directly from exploring to stable).
• Context: Governs promotions, deprecations, and exceptions across patterns, decisions, and capabilities.
• Force: Agents can hallucinate non-existent status changes or skip required verification stages
• Force: Auditability requires strict adherence to documented governance lifecycles

Traceability

Artifact source: proposed-patterns