CYBERMINDSPACE LABS / PLATFORM

The lab is only the top layer.

A proposed architecture connecting scenario intent to execution and evidence.

PROPOSED / RESEARCH

Labs should not be
static machines.

FIXED EXERCISEEXPERIMENT CONTRACT

FIXED VM

TOPOLOGY

IDENTITY

TELEMETRY

EVALUATION

REPLAY

POLICY

Explore the experiment lifecycle 30%

An explanatory diagram. Existing cyber ranges already support automation; programmability alone is not a novelty claim.

PROPOSED LIFECYCLE / 01

define.

Declare identities, assets, boundaries and the security property.

stage: define
contract: no_unauthorized_export
execution: architecture_walkthrough
production_engine: not_deployed

Conceptual cutaway of a layered computing environment and its observable data flows — RESEARCH STUDY / SYSTEM BOUNDARIES

SCENARIO-AS-CODE / RESEARCH HYPOTHESIS

What if security environments
could be compiled?

support-agent.scenario.yamlEXECUTABLE FIXTURE RULES

1schema: cms.fixture.v1
2scenario: support-agent
3mode: deterministic_browser_fixture
4
5identity: support_reader
6asset: CRM-042
7tools: [export]
8adversarial_document: false
9enforce_authorization: false
10independent_sink_witness: true
11
12property: no_unauthorized_export
13verdict: satisfied

Change the scenario.

Expose export toolIntroduce indirect injectionCapture independent sink evidence

satisfied

The complete resource witness shows no unauthorized export within this fixture’s observation window.

Authorization is deliberately disabled in this fixture. The controls run the same evaluator as the export experiment; they do not compile infrastructure.

CONTROL PLANE

Validate intent. Plan resources. Reconcile lifecycle.

Scenario registry, policy validation, scheduling and bounded retries.

EXECUTION PLANE

Run isolated, disposable environments.

An agent, synthetic services, explicit identities and controlled egress.

EVIDENCE PLANE

Observe effects independently.

Authenticated events, state witnesses, provenance and evidence retention.

EVALUATION PLANE

Compare state with the contract.

Satisfied, violated or inconclusive verdicts, with separate infrastructure status.

R&D ROADMAP / NO COMMITTED DATES

Build the foundation.
Earn the next layer.

PHASE / 01

Foundation

Basic cybersecurity and prompt-injection labs. Classroom workflows. Environment lifecycle.

IN DEVELOPMENT

PHASE / 02

Programmable environments

Scenario-as-code, dynamic provisioning, isolation and telemetry.

RESEARCH

PHASE / 03

AI-security infrastructure

Agents, RAG, MCP, tools and validated automated evaluation.

RESEARCH

PHASE / 04

Environment intelligence

Constrained scenario generation, event graphs and reproducible experiments.

LONG-TERM RESEARCH

BUILD THE RESEARCH WITH US