NexusClaw Lab / Control Models
Control models for governed execution
The lab turns NexusClaw methods into models teams can remember, discuss, and verify. Acronyms create a shared language; standard capabilities define the actual boundary.
01 / Operating models
Six models for the operating chain
These models come from the current platform positioning, growth loop, governance, developer, and A2A content. They do not introduce unverified product capabilities.
E2E | Execution Chain
End-to-end systems thinking
Connect business context, Agent actions, governance, human handoff, audit, and attribution in one production chain.
KG | Context Graph
Knowledge graphs and relationship modeling
Organize business objects, context, and permission boundaries through a shared relationship model.
HITL | Human-in-the-Loop
Human-in-the-loop control
Return high-risk, exceptional, and judgment-heavy actions to an accountable human owner.
PDCA | Growth Loop
Plan–Do–Check–Act improvement cycle
Move human corrections and real outcomes through evaluation, approval, release, and rollback before improvement.
RACI | Accountability Chain
Responsibility assignment matrix
Keep the executor, accountable owner, approver, and informed parties visible across actions and outcomes.
A2A | Open Execution
Agent-to-Agent collaboration
Bring nexus CLI, the open-source community, MCP, and A2A into one developer path.
02 / Trust models
Six capabilities, six verifiable models
Security, Compliance, Privacy, Reliability, Performance, and Scale stay as the standard names. The model labels make the discussion easier to enter.
ABAC | Attribute Boundaries
Attribute-Based Access Control
Maps to the security page’s subject, tenant, object, action, and risk decisions.
GRC | Evidence Loop
Governance, Risk, and Compliance
Maps to controls, owners, evidence, and review cycles on the compliance page.
PbD | Privacy by Design
Privacy by Design
Maps to purpose, context, model routing, and retention boundaries on the privacy page.
SRE | Recovery Engineering
Site Reliability Engineering
Maps to release, failure, failover, and recovery SOPs on the reliability page.
SLO | Performance Budgets
Service objectives and capacity budgets
Maps to latency, concurrency, capacity, and load evidence; it does not create an SLA promise.
QoS | Scale Boundaries
Quality of Service and resource allocation
Maps to tenants, queues, quotas, and deployment boundaries on the scale page.