The execution safety layer for AI agents.
-
Updated
Oct 26, 2026 - Python
The execution safety layer for AI agents.
Independent Agent Control Standard conformance testing, AI agent runtime security benchmark, adversarial side-effect validation, AgBOM, OpenTelemetry, and control economics.
VAOM Guardian: the VAOM v5 reference implementation. A fail-closed OWASP Agent Control Standard (ACS) Guardian that compiles a Delegation Authority Matrix into policy, with evidence that the boundaries hold. vaom.ai
Deterministic session-risk scoring for AI chatbots & agents: live prompt-injection / data-exfil / resource-exhaustion detection across OWASP LLM & Agentic 2026, acting via flag · redact · block · session time-out. Zero-dependency core.
Delegation Lab: an interactive VAOM simulation. Four rounds of designing AI delegation authority, enforced by the VAOM Guardian. Play at lab.vaom.ai
To associate your repository with the agent-control-standard topic, visit your repo's landing page and select "manage topics."