Tool, target, purpose, principal, and context enter one declared contract.
5D resources
Learn the boundary. Inspect the evidence.
Guides explain the implementation questions. Evidence records show what 5D supports today, what the examples demonstrate, and where the limits remain.
Implementation overview
One policy path for throughput, control, and shared evidence.
For high-volume agent applications, keep declared actions on an inspectable policy path and route only the cases your policy selects for optional semantic evidence or review.
Versioned rules resolve declared inputs on documented integration paths.
Observer evidence can tighten or hold. It cannot clear an explicit block.
Illustrative workload model
Change the declared assumptions.
Declared assumptions, not measured 5D performance.
365-day illustrative year. Currency is a display/input unit; no exchange-rate conversion is applied.
Deterministic-only actions / day6,500
Observer- or review-routed actions / day3,500
Estimated routed observer spend / year€6,387.50
One collaboration layer
Five owners, one versioned action contract.
Policy intent, implementation, threat review, human disposition, and reconstructable evidence stay connected without transferring organisational accountability to a model.
Start from public primitives: inspectable policy evaluation, verdicts, documented adapters and gateways, and evidence records. The support matrix states what version 0.7.0 verifies and what the host still owns.
Routing share depends on your policy and traffic.
Observer cost is user-supplied; it is not a 5D price or forecast.
Estimated variable observer spend excludes deterministic infrastructure, human review, retries, and other deployment costs.
Model cost and latency vary with provider, model, caching, concurrency, and deployment design.
CISO lighthouse
The AI agent governance lifecycle.
A source-linked control map across EU, UK, Singapore, NIST, ISO, and OWASP, with 5D contribution and organisational responsibility kept side by side.
Lifecycle control map
Trace authority from design through decommissioning.
See seven control stages, a separate action-attempt lifecycle, and the exact boundary between technical evidence and retained accountability.
Open the CISO map ↗ Lifecycle artifact mapSee what evidence 5D produces.
Follow five technical outputs across the lifecycle, who uses them, why they matter, and which responsibilities remain outside 5D.
Open the artifact map ↗Guides
Three implementation paths.
Each guide is matched to the tested release and reviewed against the available evidence.
Guide 01 · action boundary
Place authority before execution
An explanation matched to the tested 5D release: where the library evaluates a proposed action, what the verdict means, and what remains the host application's responsibility.
Read guide ↗Guide 02 · failure and retryKeep retries and unknown outcomes visible
How to preserve the difference between a policy decision, an attempted execution, a confirmed result, and an outcome that is still unknown.
Read guide ↗Guide 03 · decision evidenceReconstruct and challenge a decision
The minimum evidence needed to understand which action was proposed, which policy applied, how the path resolved, and where a reviewer can challenge it.
Read guide ↗Evidence
7 topics, stated with their limits.
Evidence pages separate shipped behavior, worked examples, and framework mappings. They do not establish compliance.
Authority
Who can permit an AI agent action?
Policy, observer, and reviewer roles at the action boundary.
OperationsWhat happens when an AI agent action fails, retries, or has an unknown outcome?
Phase-specific states without exactly-once claims.
PrivacyWhat data crosses the AI agent action boundary?
What crosses to the observer, stays local, and enters evidence.
ReconstructionHow can an AI agent decision be reconstructed?
A versioned schema, example receipts, stable serialization, and ordered events.
IntegrationWhere does 5D integrate into an agent runtime?
What release 0.7.0 implements, what remains partial, and what the host must own.
OversightHow does 5D map to AI governance frameworks?
Evidence concepts, primary sources, and limitations.
Research lineageWhy does 5D score five action-risk dimensions?
The founder origin, research lineage, and source boundaries behind DTREA.