Analytics Legends logoAI & Analytics Legends Sign inSign up
Academy module

Agent Memory and Runtime Safety for Joule Agents — HANA Cloud Agent Memory Service and NVIDIA OpenShell

Agent Memory and Runtime Safety for Joule Agents — HANA Cloud Agent Memory Service and NVIDIA OpenShell — AI & Analytics Legends section illustration for the SAP AI & Analytics knowledge base (concepts, studies, Academy)

As of 2026-11-07

A security-design module for Joule agents that keep state. It explains what the SAP HANA Cloud Agent Memory Service pattern does (scope derived from the authenticated user through CAP, agentID and invokerID, sharing and revocation through a separate user-confirmed skill), dates it: demonstrated on 30 September 2026, then cited in an SAP blog of 27 October 2026 (TechEd day 1); status unconfirmed (no SAP GA page found on 11 November 2026), and teaches how to design what an agent remembers, isolate and revoke it, and defend against memory poisoning (OWASP ASI06, MINJA). It then covers NVIDIA OpenShell, embedded by SAP in the Joule Studio runtime: four policy layers, YAML policy, credential injection, and the open item that links to SAP authorization, IAM and audit are under development. A ten-case red-team checklist and a hands-on lab close the module.

What you will learn

  • Write a one-page memory charter for a business agent with store, why, readers, lifetime and removal for each memory class
  • Model identity-derived memory scope with a CAP @restrict annotation and a separate, user-confirmed sharing grant that no agent tool can write
  • Write and run a revocation test that grants, uses, revokes, re-asks and proves that derived summaries and embeddings are purged
  • Draft a least-reach OpenShell sandbox policy and say which layers are fixed at creation and which can be replaced with openshell policy set
  • Score the ten-case memory and runtime red-team checklist as pass, fail or not-testable and state the owner and fix for each fail
  • Quote the status of the Agent Memory Service, OpenShell in Joule Studio and the TechEd 2026 items with date and source

Module overview

Who this is for. You can build and test a Joule agent and you know the basic security model for SAP AI (“Security for SAP AI — Prompt Injection, Data Exfiltration and Authorisations”: prompt injection, exfiltration, authorisations). This module covers the two design problems that appear the moment an agent stops being stateless: what it is allowed to remember, and what it is allowed to do at runtime whatever it has remembered. The first is addressed by the SAP HANA Cloud Agent Memory Service, the second by the NVIDIA OpenShell runtime that SAP is embedding in Joule Studio; SAP has not confirmed its availability (no release note or Help page as of 11 November 2026). Both are recent: read the status lines before you quote anything to a client. The platform vocabulary is SAP Business AI Platform; SAP HANA Cloud keeps its name. By the end of the roughly 100-minute lab in chapter 4 you can write a memory charter, a scope model, a revocation test and a least-reach runtime policy, and run a ten-case red-team checklist against them.

Source: SAP Community — Give business agents memory with SAP HANA Cloud Agent Memory Service

Prerequisites

  • Completion of Joule Agents in Production and Security for SAP AI or equivalent hands-on experience with agent testing and prompt-injection defences
  • Working knowledge of SAP CAP authorization annotations (@requires, @restrict) and of SAP HANA Cloud
  • Optional: access to the SAP Community posts of 28 to 30 September 2026 and to NVIDIA's OpenShell documentation to reproduce the lab

Outcomes

  • Describe the Agent Memory Service pattern and its status accurately to a client, without presenting a demonstration as a product commitment.
  • Produce a memory charter and an isolation design in which scope comes from identity and sharing is a separate user-confirmed grant.
  • Write a revocation test that checks recipient answers, caches and derived stores.
  • Draft a least-reach sandbox policy and explain which layers are static and which can be reloaded.
  • Run the ten-case memory and runtime red-team checklist and report not-testable cases as findings.

Full module with the Consultant plan — €19.90/month. The full module adds: the decision framework · the end-to-end scenario walkthrough · the KPI scorecard · the anti-patterns · the code blocks · the knowledge check.

Open in the app →

Open missions

67 live roles and contracts in SAP AI & Joule right now — see the SAP AI & Joule missions · freelance SAP contracts.