Objective
Define the desired security outcome, target, reduction goal, and operating constraints.
Strategy
Compare competing counterfactual action bundles rather than trusting a single opaque recommendation.
Governor
Apply policy, reliability, blast-radius, rollback, provider-safety, and approval gates before action.
The decision lifecycle
Current Security State
Canonical, truth-labelled evidence describes what is observed now: exposures, attack paths, identities, permissions, sensitive data, runtime context, crown jewels, and objective pressure.
Decision Objective
The system creates an explicit parent security intent with desired outcome, constraints, target, and reduction goal. Decisions are evaluated against that objective rather than against alert volume.
Strategy Arena
Potential remediation approaches are represented as competing action bundles. Cloudryption can compare expected security-state changes and select a recommended strategy while keeping alternatives visible.
Decision Governor
The governor determines whether execution authority is justified. Policy, confidence, reliability, blast radius, rollback readiness, provider safety, and approval requirements can block progression.
Approval & ActionContract
Only recommendations bound to registered, validated ActionContracts with complete target data can become executable Decision OS actions. Generic or incomplete recommendations remain non-executable.
Action V2 Execution
Approved remediation is handed to the canonical Action V2 worker. The execution record remains distinct from the subsequent verification record.
Post-change Evidence
Cloudryption recalculates the affected security state against the execution-bound baseline and action identity. This establishes what actually changed after execution.
Decision Proof
Predicted outcomes are compared with verified actual outcomes. Cloudryption does not manufacture after-values or treat a successful execution call as proof of security improvement.
Decision Memory
The lifecycle becomes durable, replayable history: evidence, objective, strategies, authority, approval, execution, verification, and proof.
Execution authority is earned, not assumed
Cloudryption is designed for controlled enterprise remediation. A high-ranked recommendation is not automatically an executable action. The Decision Governor and ActionContract boundary preserve the distinction between intelligence, authority, and change.
Execution is not verification
A remediation worker can report success while the intended security outcome remains unproven. Cloudryption therefore separates execution from post-change evidence and verification, allowing Decision Proof to compare predicted and actual results.
Where CID fits
CID — Cloud Intelligence Decisions — remains the intelligence and reasoning capability within Cloudryption. The customer-facing operating model is broader: Cloudryption Decision OS combines evidence, CID reasoning, decision governance, Action V2 execution, verification, proof, and memory into one controlled lifecycle.
See a complete decision lifecycle
Use a realistic cloud scenario to follow evidence through objective, strategy, governance, execution, verification, and Decision Proof.