AI security with Monofor: gateways, guardrails and machine credentials
Build the full AI security stack: a threat model for agents, v2026.11 agent budgets and approvers, the AI, MCP and A2A gateways, AI flows, shadow AI and spend control, evidence for auditors, and short-lived machine credentials in Monopam. Written for Monosign v2026.11 and Monopam v2026.09.
Monosign University Pass 1 h 7 minAdvanced DE · EN · ES · TR Administrator Security analyst Architect Partner Developer
Version 9 · updated Nov 11, 2026What changed?
Sign in to start
Included with University Pass.
What you will learn
- Map AI agent risks such as prompt injection, tool abuse and denial of wallet to concrete identity controls
- Put models, MCP tools and agent-to-agent calls behind the Monosign gateways with assignment, audience and allow-list rules
- Inspect AI traffic with AI flows, control shadow AI and spend, and produce evidence without overstating it
- Give agents short-lived machine credentials and brokered sessions in Monopam instead of standing secrets
Curriculum
4 modules · 17 lessons
Module 1: Threat model and agent governance
Module 2: The AI, MCP and A2A gateways
Module 3: Guardrails, spend and evidence
Module 4: Machine credentials for agents (Monopam)