live wire
▸AI · Red Hat documents usage-based admission fair sharing for Kueue 1.4 on OpenShiftRed Hat Developer▸AI: Red Hat maps governed firewall changes from ServiceNow through Ansible and two human approval gatesRed Hat Developer▸CLUSTER MGMT · ACM 2.17 makes Submariner 0.24 GA with Important-rated fixesRed Hat Errata▸PLATFORM · Red Hat makes on-premises Lightspeed recommendations GA for Satellite 6.18Red Hat Errata▸SECURITY · Red Hat Hardened Images updates Tomcat 10 for nine authentication, access-control and DoS flawsRed Hat Errata▸AI · Open Data Hub 3.6.0 EA1 bundles Trainer, MLflow and llm-d componentsOpen Data Hub▸AI · Speculators 0.6.0 adds P-EAGLE parallel drafting for vLLM speculative decodingRed Hat Developer▸SECURITY · OpenShift 4.17.57 fixes seven Go and TLS CVEs in an Important-rated updateRed Hat Errata▸AI · Red Hat benchmarks local LLM guardrails with EvalHub, exposing regex accuracy and latency trade-offsRed Hat Developer▸AI · Red Hat maps silent tool-call failures across agentic pipelinesRed Hat▸API · Kuadrant 1.5.3 adds GRPCRoute policies and developer-portal API-key workflowsKuadrant▸AI · (Aug 25) IBM releases Apache-2.0 Granite 4.2 reasoning models in 3B, 8B and 30B sizesIBM Research▸JAVA · Red Hat build of Quarkus 3.33.3.SP1 fixes 13 CVEs in an Important-rated updateRed Hat errata▸AI · vLLM moves Kimi K2 RL weight sync across 384 H100s in 7.53 seconds (Aug 22)vLLM▸AI · Red Hat documents usage-based admission fair sharing for Kueue 1.4 on OpenShiftRed Hat Developer▸AI: Red Hat maps governed firewall changes from ServiceNow through Ansible and two human approval gatesRed Hat Developer▸CLUSTER MGMT · ACM 2.17 makes Submariner 0.24 GA with Important-rated fixesRed Hat Errata▸PLATFORM · Red Hat makes on-premises Lightspeed recommendations GA for Satellite 6.18Red Hat Errata▸SECURITY · Red Hat Hardened Images updates Tomcat 10 for nine authentication, access-control and DoS flawsRed Hat Errata▸AI · Open Data Hub 3.6.0 EA1 bundles Trainer, MLflow and llm-d componentsOpen Data Hub▸AI · Speculators 0.6.0 adds P-EAGLE parallel drafting for vLLM speculative decodingRed Hat Developer▸SECURITY · OpenShift 4.17.57 fixes seven Go and TLS CVEs in an Important-rated updateRed Hat Errata▸AI · Red Hat benchmarks local LLM guardrails with EvalHub, exposing regex accuracy and latency trade-offsRed Hat Developer▸AI · Red Hat maps silent tool-call failures across agentic pipelinesRed Hat▸API · Kuadrant 1.5.3 adds GRPCRoute policies and developer-portal API-key workflowsKuadrant▸AI · (Aug 25) IBM releases Apache-2.0 Granite 4.2 reasoning models in 3B, 8B and 30B sizesIBM Research▸JAVA · Red Hat build of Quarkus 3.33.3.SP1 fixes 13 CVEs in an Important-rated updateRed Hat errata▸AI · vLLM moves Kimi K2 RL weight sync across 384 H100s in 7.53 seconds (Aug 22)vLLM
upstreambeat.ai
analysisFIELD BUILDS

MemoryHub packages governed shared memory for AI-agent fleets on OpenShift AI

The Red Hat AI Americas field build combines MCP access, scoped storage, provenance and audit controls in a reference architecture for teams running many agents.

OpenShift AI memory service architecture
AI-generated illustration
By The News Desk· Aug 21, 2026

MemoryHub, a field-built reference project from Red Hat AI Americas, treats AI-agent memory as a governed platform service rather than a private scratchpad attached to one assistant. The project’s repository now describes both a local edition for individual developers and a cluster edition for teams running agents on Red Hat OpenShift AI.

What the pattern changes

The project starts from a useful operational distinction: a single developer may be well served by local files or an assistant’s built-in memory, while a fleet of agents needs identity, access boundaries and a record of what information influenced an action. MemoryHub’s cluster design puts PostgreSQL and pgvector behind an MCP server, then adds scoped access control, version history, provenance branches, contradiction handling and audit events, according to the project documentation.

That makes memory a shared platform concern. Project-scoped records can be reused by agents working in the same context, while broader scopes support cross-project or organizational knowledge. The repository says every read, write, update and deletion is subject to scope enforcement, and that enterprise-scoped memories require human approval.

The OpenShift AI deployment

The cluster edition is designed for OpenShift AI and packages PostgreSQL with pgvector, MinIO object storage, OAuth 2.1 support, on-cluster embedding and reranking models, and a PatternFly-based dashboard. The README also describes air-gapped deployment, Red Hat UBI images and an installation path driven by oc and make install.

Agents reach the service through MCP, a typed Python SDK or a command-line client. The MCP layer exposes memory operations over streamable HTTP, while the SDK and CLI handle search, reads, writes and project configuration. A separate local edition uses SQLite and can be installed without cluster infrastructure.

Where platform teams should be cautious

This is a field build, not a supported Red Hat product. Its repository is active and broad, but it also marks several areas—including an operator, expanded observability and some integrations—as planned work. Teams evaluating it should treat the architecture as a reference pattern, validate the access-control and retention behavior against their own policies, and avoid assuming roadmap items are complete.

The more important contribution is the boundary it draws. Once agents share operational knowledge, retrieval quality is only one part of the problem; authorization, provenance, contradiction handling and forensic reconstruction become platform requirements. MemoryHub gives OpenShift AI teams a concrete implementation to examine rather than leaving those controls as an abstract governance checklist.

sources

Filed by The News Desk. Corrections: desk@upstreambeat.ai · Our standards →

comments · 0

    Comments are moderated before they appear. Your email is used once to confirm it is you — never shown, never sold. Corrections and questions get an answer from the desk when we have one.

    more from the desk