Agent Runtime Assurance
ConstantX turns threat-model-driven runtime tests into verifiable deployment evidence: structural containment verdicts, target-runtime traceability, and immutable evidence chains.
The Problem
Capability benchmarks (SWE-bench, MMLU) answer: "Can the model do the task?"
ConstantX answers: "When the model fails, does the target system contain it with evidence?"
For autonomous systems, safety is not about high success rates. It is about bounded failure envelopes. A system that fails safely 100% of the time is deployable (albeit useless). A system that succeeds 99% of the time but exhibits undefined behavior 1% of the time is not.
Methodology: Decision Coverage
Every autonomous run is classified into one of three verdicts:
-
valid_commit (Success)
The agent completed the task within all defined constraints. -
bounded_failure (Safe Failure)
The agent failed, but the failure was caught by the target runtime's enforcement mechanism (e.g., policy denial, step budget, sandbox block). -
undefined_behavior (Unsafe)
The agent broke the protocol, hallucinated a tool, or produced an uncaught side effect.
System Architecture
ConstantX measures agentic AI deployments across multiple target runtime architectures.
- Target enforcement surfaces: Policy denials, hard gates, verification gates, progress detection, and sandbox blocks where the target runtime exposes them.
- Signals: Observable target-runtime traces with cryptographic hashes at every gate.
- Verdict: Reduces traces to a deterministic three-state coverage outcome.
- Evidence: Packages artifacts into audit-grade evidence chains bound to dated model snapshots.
Evidence Access
Submitted to NIST AI Agent Standards Initiative (NIST-2025-0035) · Framework mappings: NIST AI RMF, OWASP ASI, MITRE ATLAS · Every adversarial scenario traces to a documented threat model entry