Whitepaper
AIMX + ILB: Governed identity and action-boundary enforcement for tool-using AI
Summary
Epydios describes two complementary runtime layers:
- Adaptive Identity Matrix Engine (AIMX): governed identity and authority as runtime state inside the loop
- Independent Layer Boundary - Behavioral Autonomy Constraint Kernel (ILB; ILB-BACK): action-boundary interposition, evaluation and enforcement where actions become effects
Together, they support a runtime story where identity and authority are explicit at decision time, enforcement is anchored at execution boundaries and evidence artifacts bind decision to outcome.
What you will find in the whitepaper
- A shared vocabulary for identity, role, authority, posture, action boundaries and evidence artifacts
- A conceptual placement model showing where AIMX and ILB sit in a tool-using AI runtime
- A runtime handshake loop that links identity and authority context to boundary evaluation and constrained execution
- Scenario illustrations and stack fit framing
- A standards alignment section framed as control mapping and evidence expectations
Table of contents
- Problem framing
- Requirements for governable autonomy
- Placement at a glance
- Vocabulary
- AIMX
- ILB
- AIMX → ILB handshake
- Benefits and behavioral shift
- Scenarios
- Fit with existing stacks
- Standards alignment and control mapping
- Limitations and non-goals
- Adoption guidance
- IP posture and diligence materials
- Next steps and feedback
- Glossary
Figures
Scope
The whitepaper is intended to disclose a conceptual model, shared vocabulary and runtime story. It is not intended to publish implementation-enabling details. To gain a better understanding contact us.