# OpenShell

NVIDIA OpenShell is an open-source runtime for executing autonomous AI agents in sandboxed environments with kernel-level isolation.

This is one page of public article previews, not the complete archive. Follow Next page to continue. Summaries are not the original full articles.

## "The best escape room players that you've ever met": What Red Hat's AI architect really thinks about agents.

DevFeed: ["The best escape room players that you've ever met": What Red Hat's AI architect really thinks about agents.](<https://devfeed.tech/articles/the-best-escape-room-players-that-you-ve-ever-met-what-red-hat-s-ai-architect-really-thinks-about-agents-141167.md>)

Original publisher: [Read original article](<https://thenewstack.io/openshell-ai-agent-security/>)

Author: Steven J. Vaughan-Nichols

Published: 2026-11-18T20:25:18Z

Content type: news

Language: en

Sources: [The New Stack](<https://devfeed.tech/sources/the-new-stack.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [redhat](<https://devfeed.tech/topics/redhat.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [agent-substrate](<https://devfeed.tech/tags/agent-substrate.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-security](<https://devfeed.tech/tags/ai-security.md>), [code](<https://devfeed.tech/tags/code.md>), [container](<https://devfeed.tech/tags/container.md>), [containers](<https://devfeed.tech/tags/containers.md>), [control](<https://devfeed.tech/tags/control.md>), [database](<https://devfeed.tech/tags/database.md>), [egress](<https://devfeed.tech/tags/egress.md>), [github](<https://devfeed.tech/tags/github.md>), [kubecon-cloudnativecon-na-2026](<https://devfeed.tech/tags/kubecon-cloudnativecon-na-2026.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

Red Hat AI architect Jessica Forrester argues that agents tasked with completing goals may drift or probe unintended resources, and that confusion can look like compromise. Her approach assumes an agent is compromised: enforce sandbox policies for network, filesystem and process access, since prompts alone are insufficient. In the demo, OpenShell allowed access to the model and GitHub; it blocked and logged disallowed destinations. Its policy advisor could propose extra network rules, but a human had to approve them by default. OpenShell also masks credentials inside the sandbox and swaps them on the way out. For stronger isolation, Red Hat runs sandboxes in Kata Containers, which give workloads lightweight VMs with hardware-backed isolation and separate Linux kernels. The article notes that these controls still depend on the underlying environment holding.

### Source excerpt

Last week at KubeCon + CloudNativeCon North America, a major gathering of the open-source and cloud-native communities, many speakers -- The post "The best escape room players that you've ever met": What Red Hat's AI architect really thinks about agents. appeared first on The New Stack.

## Securing third-party AI agents on Red Hat AI with OpenShell

DevFeed: [Securing third-party AI agents on Red Hat AI with OpenShell](<https://devfeed.tech/articles/securing-third-party-ai-agents-on-red-hat-ai-with-openshell-139445.md>)

Original publisher: [Read original article](<https://developers.redhat.com/articles/2026/11/17/omnigent-case-study-securing-third-party-ai-agents-on-red-hat-ai-with-openshell>)

Author: Sean Condon

Published: 2026-11-17T14:01:38Z

Content type: tutorial

Language: en

Sources: [Red Hat](<https://devfeed.tech/sources/red-hat.md>), [Red Hat Developer](<https://devfeed.tech/sources/red-hat-developer.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [redhat ubi](<https://devfeed.tech/topics/redhat-ubi.md>)

Tags: [agent-harnesses](<https://devfeed.tech/tags/agent-harnesses.md>), [agentops](<https://devfeed.tech/tags/agentops.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [antigravity](<https://devfeed.tech/tags/antigravity.md>), [api](<https://devfeed.tech/tags/api.md>), [api-keys](<https://devfeed.tech/tags/api-keys.md>), [apis](<https://devfeed.tech/tags/apis.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [best](<https://devfeed.tech/tags/best.md>), [best-practices](<https://devfeed.tech/tags/best-practices.md>), [bubblewrap](<https://devfeed.tech/tags/bubblewrap.md>), [building](<https://devfeed.tech/tags/building.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [red-hat](<https://devfeed.tech/tags/red-hat.md>), [sandbox](<https://devfeed.tech/tags/sandbox.md>), [security](<https://devfeed.tech/tags/security.md>), [third-party](<https://devfeed.tech/tags/third-party.md>), [third-party-ai](<https://devfeed.tech/tags/third-party-ai.md>), [third-party-ai-agents](<https://devfeed.tech/tags/third-party-ai-agents.md>)

### AI overview

Red Hat AI 3.6's bring-your-own-agent approach uses OpenShell sandboxes to run third-party agent harnesses. The excerpt demonstrates Omnigent on OpenShift: build a UBI-based sandbox image with only the needed harnesses and a restrictive base policy, configure OpenShell profiles and providers to inject credentials, then set Omnigent's OpenShift overlay to use those providers and deploy it. Credentials are masked and substituted by OpenShell's network proxy; filesystem access and outbound endpoints are blocked by default, with policy exceptions available for review and approval during a live session. Omnigent adds a shared interface for harnesses such as Claude Code, Codex, Pi, and Goose, with persistent, shareable sessions; BubbleWrap adds isolation inside the host. Its integration passes environment variables but not files, so configuration must be included in the image, increasing image maintenance. The excerpt describes this as a community proof of concept; Red Hat AI 3.6 features such as fast provisioning and native multitenancy are presented as upcoming.

### Source excerpt

The AgentOps strategy in Red Hat AI 3.6 is bring your own agent (BYOA), providing a sandbox environment for running agent h The post Securing third-party AI agents on Red Hat AI with OpenShell appeared first on Red Hat Developer.

## As AI agents take control, Nvidia rethinks the security playbook

DevFeed: [As AI agents take control, Nvidia rethinks the security playbook](<https://devfeed.tech/articles/as-ai-agents-take-control-nvidia-rethinks-the-security-playbook-131019.md>)

Original publisher: [Read original article](<https://siliconangle.com/2026/11/12/ai-networking-nvidia-scalein-security-agentic-aifactoriesdatacenters/>)

Author: Mark Albertson

Published: 2026-11-12T22:46:33Z

Content type: news

Language: en

Sources: [SiliconANGLE](<https://devfeed.tech/sources/siliconangle.md>)

Topics: [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [ai security](<https://devfeed.tech/topics/ai-security.md>), [Security](<https://devfeed.tech/topics/security.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [BlueField DPU](<https://devfeed.tech/topics/bluefield-dpu.md>), [DOCA](<https://devfeed.tech/topics/doca.md>), [observability](<https://devfeed.tech/topics/observability.md>), [NVIDIA Spectrum-X Ethernet](<https://devfeed.tech/topics/nvidia-spectrum-x-ethernet.md>), [tenant data protection](<https://devfeed.tech/topics/tenant-data-protection.md>), [Access Control](<https://devfeed.tech/topics/access-control.md>), [NVIDIA Vera Rubin](<https://devfeed.tech/topics/nvidia-vera-rubin.md>), [Nvidia](<https://devfeed.tech/topics/nvidia.md>)

Tags: [agent-safety](<https://devfeed.tech/tags/agent-safety.md>), [agent-security](<https://devfeed.tech/tags/agent-security.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agentic-ai-security](<https://devfeed.tech/tags/agentic-ai-security.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-factories](<https://devfeed.tech/tags/ai-factories.md>), [ai-factory-roi-monetization-flywheels](<https://devfeed.tech/tags/ai-factory-roi-monetization-flywheels.md>), [ai-infrastructure-finance](<https://devfeed.tech/tags/ai-infrastructure-finance.md>), [ai-networking](<https://devfeed.tech/tags/ai-networking.md>), [aifactoriesdatacenters](<https://devfeed.tech/tags/aifactoriesdatacenters.md>), [aifactoriesdatacenters25eventpage](<https://devfeed.tech/tags/aifactoriesdatacenters25eventpage.md>), [bluefield-4](<https://devfeed.tech/tags/bluefield-4.md>), [capex-trends](<https://devfeed.tech/tags/capex-trends.md>), [cube-event-coverage](<https://devfeed.tech/tags/cube-event-coverage.md>), [data-centers-of-the-future](<https://devfeed.tech/tags/data-centers-of-the-future.md>), [doca](<https://devfeed.tech/tags/doca.md>), [dpu](<https://devfeed.tech/tags/dpu.md>), [gilad-shainer](<https://devfeed.tech/tags/gilad-shainer.md>), [gpu-access](<https://devfeed.tech/tags/gpu-access.md>), [hybrid-cloud-ecosystems](<https://devfeed.tech/tags/hybrid-cloud-ecosystems.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [isolation](<https://devfeed.tech/tags/isolation.md>), [nasdaq-nvda](<https://devfeed.tech/tags/nasdaq-nvda.md>), [news](<https://devfeed.tech/tags/news.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [orchestration-and-automation](<https://devfeed.tech/tags/orchestration-and-automation.md>), [scale-in](<https://devfeed.tech/tags/scale-in.md>), [secure-access](<https://devfeed.tech/tags/secure-access.md>), [security](<https://devfeed.tech/tags/security.md>), [thecube](<https://devfeed.tech/tags/thecube.md>), [thecube-plus-nyse-wired-ai-factories-data-centers-of-the-future-2025](<https://devfeed.tech/tags/thecube-plus-nyse-wired-ai-factories-data-centers-of-the-future-2025.md>), [unit-economics](<https://devfeed.tech/tags/unit-economics.md>)

### AI overview

Nvidia's Scale-In combines BlueField-4, DOCA and Spectrum-X Ethernet to secure and operate agentic AI factories. BlueField-4's in-silicon engines collect telemetry out of band from memory, storage, GPUs and CPUs, giving agents no control over that monitoring; DOCA APIs expose the DPU's security and storage acceleration capabilities. Nvidia's OpenShell runtime governs agent execution, access and inference routing. The later OpenShell Reference Architecture adds Sentry on BlueField-4, which Nvidia says continuously monitors agent behavior and can quarantine agents that cross defined software boundaries. The architecture pairs runtime controls with an independent monitoring and enforcement layer; the excerpt describes Nvidia's design and claims, not independent performance results.

### Source excerpt

Nvidia Corp. recently added a new pillar to its AI networking infrastructure with the unveiling of Scale-In. Scale-In brings purpose-built acceleration to secure, manage and operate agentic AI factories. It is designed to deliver secure access, high-performance data movement, tenant isolation, simpler operations and more predictable performance as agents assume increasingly larger roles within AI [...] The post As AI agents take control, Nvidia rethinks the security playbook appeared first on SiliconANGLE.

## Check a subagent's OpenShell policy against a boundary

DevFeed: [Check a subagent's OpenShell policy against a boundary](<https://devfeed.tech/articles/check-a-subagent-s-openshell-policy-against-a-boundary-129766.md>)

Original publisher: [Read original article](<https://nirmata.com/2026/11/12/openshell-policy-ceilings-and-least-privilege-delegation/>)

Author: Ritesh Patel

Published: 2026-11-12T04:23:18.277431Z

Content type: tutorial

Language: en

Sources: [Nirmata](<https://devfeed.tech/sources/nirmata.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [policy-as-code](<https://devfeed.tech/topics/policy-as-code.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent-governance](<https://devfeed.tech/tags/ai-agent-governance.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-governance](<https://devfeed.tech/tags/ai-governance.md>), [allows](<https://devfeed.tech/tags/allows.md>), [boundary](<https://devfeed.tech/tags/boundary.md>), [candidate](<https://devfeed.tech/tags/candidate.md>), [coverage](<https://devfeed.tech/tags/coverage.md>), [delegation](<https://devfeed.tech/tags/delegation.md>), [docs](<https://devfeed.tech/tags/docs.md>), [domains](<https://devfeed.tech/tags/domains.md>), [effective](<https://devfeed.tech/tags/effective.md>), [filesystem](<https://devfeed.tech/tags/filesystem.md>), [gateway](<https://devfeed.tech/tags/gateway.md>), [global](<https://devfeed.tech/tags/global.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [kubernetes-security](<https://devfeed.tech/tags/kubernetes-security.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [parent](<https://devfeed.tech/tags/parent.md>), [policy](<https://devfeed.tech/tags/policy.md>), [policy-as-code](<https://devfeed.tech/tags/policy-as-code.md>), [run](<https://devfeed.tech/tags/run.md>), [sandbox](<https://devfeed.tech/tags/sandbox.md>), [subagent](<https://devfeed.tech/tags/subagent.md>)

### AI overview

OpenShell's Policy Prover checks offline whether a candidate policy stays within a maximum-access boundary; only `within_boundary` passes, and the check does not apply or enforce the policy. For a subagent, save the parent's allowed policy as the boundary and the subagent policy as the candidate, run `openshell-prover check ... --boundary ... --output json`, and inspect coverage. Use the sandbox's effective policy as the candidate when checking a live sandbox, including provider-added rules. Treat `unsupported` and `inconclusive` as evidence gaps: limitations include audit-mode endpoints, several protocols and query-parameter rules, path and binary-glob comparisons, unresolved hostnames, and resource limits. The policy advisor separately allows network-rule proposals, with manual approval by default; its risk check is distinct from the boundary check. A global policy replaces sandbox policies rather than constraining them. A pass does not establish task suitability, runtime enforcement, or Kubernetes-layer safety.

### Source excerpt

OpenShell's docs call a maximum-access policy a boundary. The prover checks that a subagent's policy allows nothing beyond the parent's. Here is how to run it.

## Why Nvidia gave OpenShell to the CNCF

DevFeed: [Why Nvidia gave OpenShell to the CNCF](<https://devfeed.tech/articles/why-nvidia-gave-openshell-to-the-cncf-129504.md>)

Original publisher: [Read original article](<https://thenewstack.io/nvidia-donates-openshell-cncf/>)

Author: Frederic Lardinois

Published: 2026-11-11T19:13:54Z

Content type: news

Language: en

Sources: [The New Stack](<https://devfeed.tech/sources/the-new-stack.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Nvidia](<https://devfeed.tech/topics/nvidia.md>), [linux foundation](<https://devfeed.tech/topics/linux-foundation.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>), [open-source-security](<https://devfeed.tech/topics/open-source-security.md>), [observability](<https://devfeed.tech/topics/observability.md>), [OpenTelemetry](<https://devfeed.tech/topics/opentelemetry.md>)

Tags: [a2a-protocol](<https://devfeed.tech/tags/a2a-protocol.md>), [agent-substrate](<https://devfeed.tech/tags/agent-substrate.md>), [agentgateway](<https://devfeed.tech/tags/agentgateway.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agentic-ai-foundation](<https://devfeed.tech/tags/agentic-ai-foundation.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-workloads](<https://devfeed.tech/tags/ai-workloads.md>), [alibaba](<https://devfeed.tech/tags/alibaba.md>), [canonical](<https://devfeed.tech/tags/canonical.md>), [cloud-native](<https://devfeed.tech/tags/cloud-native.md>), [cloud-native-computing-foundation](<https://devfeed.tech/tags/cloud-native-computing-foundation.md>), [cloud-native-ecosystem](<https://devfeed.tech/tags/cloud-native-ecosystem.md>), [cncf](<https://devfeed.tech/tags/cncf.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [security](<https://devfeed.tech/tags/security.md>), [security-governance](<https://devfeed.tech/tags/security-governance.md>)

### AI overview

Nvidia donated OpenShell to the CNCF because it sees agent security as infrastructure for the Kubernetes and cloud native teams that operate production workloads. OpenShell runs agents in sandboxes and enforces policies on file, network, credential, and model endpoint access; it is adopting CNCF project Cedar as its policy language. The CNCF offers adjacent infrastructure projects and governance, while the Linux Foundation's Agentic AI Foundation focuses on agent protocols such as MCP and A2A, and the PyTorch Foundation on AI engineering tools. OpenShell is a CNCF Sandbox project, and the team is still considering an independent security committee for vulnerability disclosures. Nvidia hopes community contributions will build integrations and tooling for compute, credentials, inference, policy, security, privacy, and observability. The agent stack remains unsettled; the article describes expected experimentation over the next six to 12 months, including integration work with Agent Substrate.

### Source excerpt

Nvidia this week donated OpenShell, its open-source runtime for securing and governing AI agents, to the Cloud Native Computing Foundation. The post Why Nvidia gave OpenShell to the CNCF appeared first on The New Stack.

## Red Hat and Nvidia's OpenShell puts AI agents on a deny-by-default leash

DevFeed: [Red Hat and Nvidia's OpenShell puts AI agents on a deny-by-default leash](<https://devfeed.tech/articles/red-hat-and-nvidia-s-openshell-puts-ai-agents-on-a-deny-by-default-leash-129342.md>)

Original publisher: [Read original article](<https://siliconangle.com/2026/11/11/prompt-injection-attacks-drive-red-hat-nvidia-s-openshell-kubeconna/>)

Author: Ryan Stevens

Published: 2026-11-11T18:55:13Z

Content type: news

Language: en

Sources: [SiliconANGLE](<https://devfeed.tech/sources/siliconangle.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [ai security](<https://devfeed.tech/topics/ai-security.md>), [redhat](<https://devfeed.tech/topics/redhat.md>), [Nvidia](<https://devfeed.tech/topics/nvidia.md>)

Tags: [agent-security](<https://devfeed.tech/tags/agent-security.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [agents-in-production](<https://devfeed.tech/tags/agents-in-production.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-agent-security](<https://devfeed.tech/tags/ai-agent-security.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-inference](<https://devfeed.tech/tags/ai-inference.md>), [ai-infrastructure](<https://devfeed.tech/tags/ai-infrastructure.md>), [attacks](<https://devfeed.tech/tags/attacks.md>), [cloud-native](<https://devfeed.tech/tags/cloud-native.md>), [cloud-native-computing-foundation](<https://devfeed.tech/tags/cloud-native-computing-foundation.md>), [cloud-security](<https://devfeed.tech/tags/cloud-security.md>), [cloudnativecon](<https://devfeed.tech/tags/cloudnativecon.md>), [cncf](<https://devfeed.tech/tags/cncf.md>), [credentials](<https://devfeed.tech/tags/credentials.md>), [cube-event-coverage](<https://devfeed.tech/tags/cube-event-coverage.md>), [cybersecurity](<https://devfeed.tech/tags/cybersecurity.md>), [data-exfiltration](<https://devfeed.tech/tags/data-exfiltration.md>), [derek-carr](<https://devfeed.tech/tags/derek-carr.md>), [developer-platforms](<https://devfeed.tech/tags/developer-platforms.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [enterprise-ai](<https://devfeed.tech/tags/enterprise-ai.md>), [gpu-infrastructure](<https://devfeed.tech/tags/gpu-infrastructure.md>), [guardrails](<https://devfeed.tech/tags/guardrails.md>), [john-furrier](<https://devfeed.tech/tags/john-furrier.md>), [kubecon](<https://devfeed.tech/tags/kubecon.md>), [kubecon-plus-cloudnativecon-na-2026](<https://devfeed.tech/tags/kubecon-plus-cloudnativecon-na-2026.md>), [kubeconna](<https://devfeed.tech/tags/kubeconna.md>), [kubeconna26eventpage](<https://devfeed.tech/tags/kubeconna26eventpage.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [mrunal-patel](<https://devfeed.tech/tags/mrunal-patel.md>), [news](<https://devfeed.tech/tags/news.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [nyse-ibm](<https://devfeed.tech/tags/nyse-ibm.md>), [observability](<https://devfeed.tech/tags/observability.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [platform-engineering](<https://devfeed.tech/tags/platform-engineering.md>), [prompt-injection](<https://devfeed.tech/tags/prompt-injection.md>), [rebecca-knight](<https://devfeed.tech/tags/rebecca-knight.md>), [red-hat](<https://devfeed.tech/tags/red-hat.md>), [security](<https://devfeed.tech/tags/security.md>), [thecube](<https://devfeed.tech/tags/thecube.md>)

### AI overview

OpenShell is an open-source project from Red Hat and Nvidia intended to help enterprises run AI agents with safeguards against prompt injection and data exfiltration. Operators can permit access to specific endpoints while keeping credentials hidden from agents and scoped to those endpoints; networking is deny-by-default. A policy advisor asks agents to explain blocked actions in plain language, and operators can update policies as workloads change. The project is under Cloud Native Computing Foundation governance following Nvidia's contribution. Red Hat aims to make deploying agents on OpenShift routine, but the excerpt offers no implementation details or comparative evaluation of the safeguards.

### Source excerpt

Prompt injection attacks and silent data exfiltration have turned AI agent security from a theoretical worry into an urgent enterprise concern. As agents move from developer laptops into data centers, companies need guardrails that don't strip away what makes agents useful. Red Hat Inc. and Nvidia Corp. have expanded their partnership to tackle that problem [...] The post Red Hat and Nvidia's OpenShell puts AI agents on a deny-by-default leash appeared first on SiliconANGLE.

## How to write OpenShell policies

DevFeed: [How to write OpenShell policies](<https://devfeed.tech/articles/how-to-write-openshell-policies-127502.md>)

Original publisher: [Read original article](<https://nirmata.com/2026/11/11/how-to-write-openshell-policies/>)

Author: Ritesh Patel

Published: 2026-11-10T23:14:06.012746Z

Content type: tutorial

Language: en

Sources: [Nirmata](<https://devfeed.tech/sources/nirmata.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Access Control](<https://devfeed.tech/topics/access-control.md>), [policy-as-code](<https://devfeed.tech/topics/policy-as-code.md>), [Documentation](<https://devfeed.tech/topics/documentation.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [filesystem](<https://devfeed.tech/tags/filesystem.md>), [how-to](<https://devfeed.tech/tags/how-to.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [kubernetes-security](<https://devfeed.tech/tags/kubernetes-security.md>), [live](<https://devfeed.tech/tags/live.md>), [network](<https://devfeed.tech/tags/network.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [policies](<https://devfeed.tech/tags/policies.md>), [policy](<https://devfeed.tech/tags/policy.md>), [policy-as-code](<https://devfeed.tech/tags/policy-as-code.md>), [rules](<https://devfeed.tech/tags/rules.md>), [sandbox](<https://devfeed.tech/tags/sandbox.md>), [verify](<https://devfeed.tech/tags/verify.md>), [write](<https://devfeed.tech/tags/write.md>), [yaml](<https://devfeed.tech/tags/yaml.md>)

### AI overview

OpenShell policies deny unlisted access and split settings by when they take effect: filesystem, Landlock, and process settings are fixed at sandbox startup, while network rules can be updated live. Write narrow network rules per binary and endpoint, naming the executable that opens the connection; rules combine every listed binary with every endpoint, and child processes inherit them. Network enforcement defaults to audit, which logs violations but allows requests; set enforce to block, and remember matching deny rules always take precedence. Landlock best_effort logs a high-severity finding and runs without filesystem rules if they cannot be enforced; hard_requirement prevents startup. After updates, check that the latest revision is Loaded and inspect the effective policy. Filesystem changes require recreating the sandbox, which stops processes and removes state.

### Source excerpt

Write an OpenShell sandbox policy in YAML: set filesystem and Landlock rules at startup, scope each network rule to one binary, then verify changes live.

## OpenShell vs Agent Substrate vs Kubernetes Agent Sandbox

DevFeed: [OpenShell vs Agent Substrate vs Kubernetes Agent Sandbox](<https://devfeed.tech/articles/openshell-vs-agent-substrate-vs-kubernetes-agent-sandbox-124302.md>)

Original publisher: [Read original article](<https://nirmata.com/2026/11/10/openshell-vs-agent-substrate-vs-kubernetes-agent-sandbox/>)

Author: Ritesh Patel

Published: 2026-11-10T00:45:38.735945Z

Content type: comparison

Language: en

Sources: [Nirmata](<https://devfeed.tech/sources/nirmata.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [AI Sandbox](<https://devfeed.tech/topics/ai-sandbox.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-sandbox](<https://devfeed.tech/tags/agent-sandbox.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [compare](<https://devfeed.tech/tags/compare.md>), [containers](<https://devfeed.tech/tags/containers.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [kubernetes-security](<https://devfeed.tech/tags/kubernetes-security.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [requires](<https://devfeed.tech/tags/requires.md>), [sandbox](<https://devfeed.tech/tags/sandbox.md>), [substrate](<https://devfeed.tech/tags/substrate.md>), [vs](<https://devfeed.tech/tags/vs.md>)

### AI overview

The three projects manage different boundaries: OpenShell governs agent access with kernel controls and declarative policy, including per-request network and credential checks through a supervisor; it supports Docker, Podman, Kubernetes, and VM runtimes. Agent Substrate runs stateful actors on pre-warmed workers, checkpointing idle actors to object storage and restoring them on demand; its overview claims restore time under 500 ms, though its README says 1.0 while its development section calls it very young. Kubernetes Agent Sandbox provides stateful Sandbox pods with stable identity and persistent storage, plus warm pools and a Python client; the cited pages do not describe checkpointing. Its basic quickstart lacks container isolation unless gVisor or Kata is configured. The excerpt gives different Kubernetes requirements for Agent Sandbox and notes the referenced Gateway API CRDs need a newer version. Choose based on the control needed, then verify release, isolation, network policy, and state handling for the version deployed.

### Source excerpt

OpenShell, Agent Substrate and Kubernetes Agent Sandbox solve different parts of running AI agents. Compare what each manages, enforces and requires.

## Securing AI Agents with NVIDIA OpenShell and Harness

DevFeed: [Securing AI Agents with NVIDIA OpenShell and Harness](<https://devfeed.tech/articles/securing-ai-agents-with-nvidia-openshell-and-harness-132619.md>)

Original publisher: [Read original article](<https://www.harness.io/blog/securing-agents-and-their-apis-with-harness-and-nvidias-openshell-architecture>)

Author: Sudhir Patamsetti

Published: 2026-11-10T00:00:00Z

Content type: article

Language: en

Sources: [Harness](<https://devfeed.tech/sources/harness-blog.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [threat detection](<https://devfeed.tech/topics/threat-detection.md>), [Threat monitoring](<https://devfeed.tech/topics/threat-monitoring.md>), [Incident response](<https://devfeed.tech/topics/incident-response.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-discovery](<https://devfeed.tech/tags/agent-discovery.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-red-teaming](<https://devfeed.tech/tags/ai-red-teaming.md>), [api](<https://devfeed.tech/tags/api.md>), [api-security](<https://devfeed.tech/tags/api-security.md>), [detection](<https://devfeed.tech/tags/detection.md>), [detection-and-response](<https://devfeed.tech/tags/detection-and-response.md>), [discovery](<https://devfeed.tech/tags/discovery.md>), [harness](<https://devfeed.tech/tags/harness.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [nvidia-openshell](<https://devfeed.tech/tags/nvidia-openshell.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [red-teaming](<https://devfeed.tech/tags/red-teaming.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [runtime-threat-detection](<https://devfeed.tech/tags/runtime-threat-detection.md>), [securing-ai](<https://devfeed.tech/tags/securing-ai.md>), [securing-ai-agents](<https://devfeed.tech/tags/securing-ai-agents.md>)

### AI overview

NVIDIA OpenShell governs agent runtime access, with optional NVIDIA Sentry on BlueField providing an independent monitoring and enforcement layer. Harness complements these controls by discovering agents through runtime traffic and mapping their connections to models, MCP servers, tools, and APIs, including observed data flows and security issues. Its red teaming tests whether malicious inputs can manipulate agents, misuse tools, or expose data; API security testing checks connected services for authentication, authorization, and data exposure weaknesses. At supported gateways and application enforcement points, Harness monitors for threats and policy violations and can apply protection against prompt injection, tool misuse, data exposure, and API attacks. The proposed deployment keeps the Harness platform, data lake, decision services, and testing runner in the customer environment; customer supplied LLMs provide local inference, using GPUs where required. The excerpt presents Rubin GPUs as a potential home for those models and Sentry as optional.

### Source excerpt

Discover how Harness complements NVIDIA OpenShell with agent discovery, AI red teaming, API security, and runtime threat detection and response. | Blog

## NVIDIA Ships Dynamo Recipes for A.X K2, K-EXAONE, Solar Open 2, and Motif-3 on B200, Claiming Up to 2x

DevFeed: [NVIDIA Ships Dynamo Recipes for A.X K2, K-EXAONE, Solar Open 2, and Motif-3 on B200, Claiming Up to 2x](<https://devfeed.tech/articles/nvidia-ships-dynamo-recipes-for-a-x-k2-k-exaone-solar-open-2-and-motif-3-on-b200-claiming-up-to-2x-124325.md>)

Original publisher: [Read original article](<https://www.storagereview.com/news/nvidia-dynamo-recipes-korean-open-models-b200-2x>)

Author: Harold Fritts

Published: 2026-11-09T16:08:45Z

Content type: news

Language: en

Sources: [StorageReview.com](<https://devfeed.tech/sources/storagereview-com.md>)

Topics: [Dynamo](<https://devfeed.tech/topics/dynamo.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [NemoClaw](<https://devfeed.tech/topics/nemoclaw.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [ai-infrastructure](<https://devfeed.tech/tags/ai-infrastructure.md>), [ai-research](<https://devfeed.tech/tags/ai-research.md>), [ai-security](<https://devfeed.tech/tags/ai-security.md>), [certification](<https://devfeed.tech/tags/certification.md>), [collaboration](<https://devfeed.tech/tags/collaboration.md>), [deployment](<https://devfeed.tech/tags/deployment.md>), [dynamo](<https://devfeed.tech/tags/dynamo.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [gpus](<https://devfeed.tech/tags/gpus.md>), [korean](<https://devfeed.tech/tags/korean.md>), [lg](<https://devfeed.tech/tags/lg.md>), [models](<https://devfeed.tech/tags/models.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [nvidia-b200](<https://devfeed.tech/tags/nvidia-b200.md>), [open](<https://devfeed.tech/tags/open.md>), [performance](<https://devfeed.tech/tags/performance.md>), [sk-telecom](<https://devfeed.tech/tags/sk-telecom.md>), [solar](<https://devfeed.tech/tags/solar.md>), [technologies](<https://devfeed.tech/tags/technologies.md>), [up](<https://devfeed.tech/tags/up.md>)

### AI overview

NVIDIA released Dynamo deployment recipes for SK Telecom's A.X K2, LG AI Research's K-EXAONE 2.0, Upstage's Solar Open 2, and Motif Technologies' Motif-3, tuned for B200 GPUs. Dynamo is an open-source distributed inference-serving framework for multi-node deployments; each recipe packages serving configuration for a specific model. NVIDIA says the recipes offer up to 2x performance out of the box, but the text gives no baseline, workload, or serving configuration for that figure. The recipes are intended to help put the models into production faster, including for services supported by the government's AI for All initiative. The text separately says LG's API access announcement names K-EXAONE-2.5, while the Dynamo recipe targets K-EXAONE 2.0.

### Source excerpt

NVIDIA released Dynamo deployment recipes for four Korean open models at AI Day Seoul: SK Telecom's A.X K2, LG AI Research's K-EXAONE 2.0, Upstage's Solar Open 2, and Motif Technologies' Motif-3. The recipes are tuned for NVIDIA B200 GPUs, and NVIDIA says they deliver up to 2x performance out of the box without naming the The post NVIDIA Ships Dynamo Recipes for A.X K2, K-EXAONE, Solar Open 2, and Motif-3 on B200, Claiming Up to 2x appeared first on StorageReview.com.

## Korea's AI Ecosystem Advances and Secures Agentic AI With NVIDIA

DevFeed: [Korea's AI Ecosystem Advances and Secures Agentic AI With NVIDIA](<https://devfeed.tech/articles/korea-s-ai-ecosystem-advances-and-secures-agentic-ai-with-nvidia-123630.md>)

Original publisher: [Read original article](<https://blogs.nvidia.com/blog/korea-secure-agentic-ai/>)

Author: Nicola Sessions

Published: 2026-11-09T05:00:39Z

Content type: news

Language: en

Sources: [NVIDIA](<https://devfeed.tech/sources/nvidia-blog.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Nemotron](<https://devfeed.tech/topics/nemotron.md>), [Agentic AI](<https://devfeed.tech/topics/what-is-agentic-ai.md>), [NemoClaw](<https://devfeed.tech/topics/nemoclaw.md>)

Tags: [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [blueprints](<https://devfeed.tech/tags/blueprints.md>), [employee](<https://devfeed.tech/tags/employee.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [hyundai](<https://devfeed.tech/tags/hyundai.md>), [intelligence](<https://devfeed.tech/tags/intelligence.md>), [korea](<https://devfeed.tech/tags/korea.md>), [models](<https://devfeed.tech/tags/models.md>), [nemoclaw](<https://devfeed.tech/tags/nemoclaw.md>), [nemotron](<https://devfeed.tech/tags/nemotron.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [nvidia-openshell](<https://devfeed.tech/tags/nvidia-openshell.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [platform](<https://devfeed.tech/tags/platform.md>), [productivity](<https://devfeed.tech/tags/productivity.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [secure-ai](<https://devfeed.tech/tags/secure-ai.md>), [secure-ai-agents](<https://devfeed.tech/tags/secure-ai-agents.md>), [security](<https://devfeed.tech/tags/security.md>)

### AI overview

NVIDIA OpenShell provides sandboxed execution and policy-based controls for enterprise AI agents. AIM Intelligence is integrating it with Starfort, whose detection of malicious instructions and data leakage is intended to complement OpenShell's runtime access controls; the company is also considering a preliminary public-sector energy proof of concept. Hyundai Motor Group is exploring NVIDIA NemoClaw blueprints and OpenShell. Separately, LG AI Research is opening API access to K-EXAONE-2.5, developed using NVIDIA Nemotron datasets, including Nemotron-Personas-Korea, a synthetic dataset of 7 million Korean personas grounded in official statistics. The excerpt describes these efforts but does not report proof-of-concept results.

### Source excerpt

As AI agents take on more work, they need capable open models and security that advances alongside them. Hyundai Motor Group is exploring NVIDIA NemoClaw blueprints and the NVIDIA OpenShell runtime, while AIM Intelligence is integrating its Starfort platform with NVIDIA OpenShell to support secure AI agents for employee productivity and enterprise workflows. A key [...]

## How NVIDIA Uses AI Agents to Streamline AI Factory Operations

DevFeed: [How NVIDIA Uses AI Agents to Streamline AI Factory Operations](<https://devfeed.tech/articles/how-nvidia-uses-ai-agents-to-streamline-ai-factory-operations-117988.md>)

Original publisher: [Read original article](<https://developer.nvidia.com/blog/how-nvidia-uses-ai-agents-to-streamline-ai-factory-operations/>)

Author: Shara Tibken

Published: 2026-11-04T22:00:17Z

Content type: article

Language: en

Sources: [NVIDIA Developer](<https://devfeed.tech/sources/nvidia-developer.md>), [NVIDIA Technical Blog](<https://devfeed.tech/sources/nvidia-technical-blog.md>)

Topics: [AI in IT operations](<https://devfeed.tech/topics/ai-in-it-operations.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [Nvidia](<https://devfeed.tech/topics/nvidia.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [AIOps for Kubernetes](<https://devfeed.tech/topics/aiops-for-kubernetes.md>), [control-plane](<https://devfeed.tech/topics/control-plane.md>), [NemoClaw](<https://devfeed.tech/topics/nemoclaw.md>), [Production Engineering](<https://devfeed.tech/topics/production-engineering.md>), [Nemotron](<https://devfeed.tech/topics/nemotron.md>), [Grafana](<https://devfeed.tech/topics/grafana.md>), [digital sovereignty](<https://devfeed.tech/topics/digital-sovereignty.md>), [DGX Spark](<https://devfeed.tech/topics/dgx-spark.md>)

Tags: [access-control](<https://devfeed.tech/tags/access-control.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agentic-ai-generative-ai](<https://devfeed.tech/tags/agentic-ai-generative-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-factory](<https://devfeed.tech/tags/ai-factory.md>), [build-ai-agents](<https://devfeed.tech/tags/build-ai-agents.md>), [clusters](<https://devfeed.tech/tags/clusters.md>), [compute](<https://devfeed.tech/tags/compute.md>), [dashboards](<https://devfeed.tech/tags/dashboards.md>), [data-center-cloud](<https://devfeed.tech/tags/data-center-cloud.md>), [dgx](<https://devfeed.tech/tags/dgx.md>), [engineers](<https://devfeed.tech/tags/engineers.md>), [gb200](<https://devfeed.tech/tags/gb200.md>), [gb300](<https://devfeed.tech/tags/gb300.md>), [gpu](<https://devfeed.tech/tags/gpu.md>), [gpu-clusters](<https://devfeed.tech/tags/gpu-clusters.md>), [hardware](<https://devfeed.tech/tags/hardware.md>), [hgx](<https://devfeed.tech/tags/hgx.md>), [hpc-scientific-computing](<https://devfeed.tech/tags/hpc-scientific-computing.md>), [machine-learning-artificial-intelligence](<https://devfeed.tech/tags/machine-learning-artificial-intelligence.md>), [mlops](<https://devfeed.tech/tags/mlops.md>), [nemoclaw](<https://devfeed.tech/tags/nemoclaw.md>), [nemotron](<https://devfeed.tech/tags/nemotron.md>), [network](<https://devfeed.tech/tags/network.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [operations](<https://devfeed.tech/tags/operations.md>), [reliability](<https://devfeed.tech/tags/reliability.md>), [scale](<https://devfeed.tech/tags/scale.md>), [scheduling](<https://devfeed.tech/tags/scheduling.md>), [site-reliability](<https://devfeed.tech/tags/site-reliability.md>)

### AI overview

NVIDIA's AI Factory Operations Agent correlates signals across compute, network, hardware, and scheduling so SREs can investigate failures through a chat interface. Specialized agents detect issues, analyze cross-layer causes, and recommend actions; examples include job recovery, hardware triage, observability, documentation search, Kubernetes awareness, and cluster management. The framework uses MCP tools and can connect to existing infrastructure integrations. It runs as a Kubernetes control-plane application, with OpenShell creating per-session sandboxes; Kubernetes inspection uses a separate read-only tool with scoped credentials, and the excerpt's example permissions deny access to secrets and pod creation or updates. NVIDIA emphasizes clear agent boundaries, read-only investigation, precise tools, and visible evidence and tool calls. A dedicated GPU-accelerated control plane can isolate orchestration from production workloads, while tenant-specific permissions and sandboxes limit visibility. The framework is experimental open source software, so APIs and deployment requirements may change.

### Source excerpt

As GPU clusters scale, operations span compute, network, hardware, and scheduling. Site reliability engineers (SREs) often have to examine multiple dashboards...

## How NVIDIA Factory Operations Blueprint Coordinates AI Agents to Investigate Production Issues

DevFeed: [How NVIDIA Factory Operations Blueprint Coordinates AI Agents to Investigate Production Issues](<https://devfeed.tech/articles/how-nvidia-factory-operations-blueprint-coordinates-ai-agents-to-investigate-production-issues-117641.md>)

Original publisher: [Read original article](<https://developer.nvidia.com/blog/how-nvidia-factory-operations-blueprint-coordinates-ai-agents-to-investigate-production-issues/>)

Author: Samuel Ochoa

Published: 2026-11-04T17:00:00Z

Content type: article

Language: en

Sources: [NVIDIA Developer](<https://devfeed.tech/sources/nvidia-developer.md>), [NVIDIA Technical Blog](<https://devfeed.tech/sources/nvidia-technical-blog.md>)

Topics: [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [agentic workflows](<https://devfeed.tech/topics/agentic-workflows.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [human in the loop agent](<https://devfeed.tech/topics/human-in-the-loop-agent.md>), [Local AI](<https://devfeed.tech/topics/local-ai.md>), [NemoClaw](<https://devfeed.tech/topics/nemoclaw.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Image processing](<https://devfeed.tech/topics/image-processing.md>), [Nemotron](<https://devfeed.tech/topics/nemotron.md>), [Isaac](<https://devfeed.tech/topics/isaac.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [2026](<https://devfeed.tech/tags/2026.md>), [3](<https://devfeed.tech/tags/3.md>), [about](<https://devfeed.tech/tags/about.md>), [actions](<https://devfeed.tech/tags/actions.md>), [adapters](<https://devfeed.tech/tags/adapters.md>), [agent-toolkit](<https://devfeed.tech/tags/agent-toolkit.md>), [agentic-ai-generative-ai](<https://devfeed.tech/tags/agentic-ai-generative-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [blueprint](<https://devfeed.tech/tags/blueprint.md>), [build-ai-agents](<https://devfeed.tech/tags/build-ai-agents.md>), [computer-vision-video-analytics](<https://devfeed.tech/tags/computer-vision-video-analytics.md>), [costs](<https://devfeed.tech/tags/costs.md>), [downtime](<https://devfeed.tech/tags/downtime.md>), [edge-computing](<https://devfeed.tech/tags/edge-computing.md>), [factory](<https://devfeed.tech/tags/factory.md>), [failed](<https://devfeed.tech/tags/failed.md>), [featured](<https://devfeed.tech/tags/featured.md>), [large](<https://devfeed.tech/tags/large.md>), [llms](<https://devfeed.tech/tags/llms.md>), [nemo](<https://devfeed.tech/tags/nemo.md>), [nemoclaw](<https://devfeed.tech/tags/nemoclaw.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [operations](<https://devfeed.tech/tags/operations.md>), [production](<https://devfeed.tech/tags/production.md>), [report](<https://devfeed.tech/tags/report.md>), [robotics](<https://devfeed.tech/tags/robotics.md>), [siemens](<https://devfeed.tech/tags/siemens.md>)

### AI overview

NVIDIA Factory Operations Blueprint (FOX) is a reference architecture for a factory manager agent that connects existing factory systems and delegates work to specialist agents. In the SMT example, an AOI alert prompts a second image inspection; confirmed defects can be sent for rework while a root cause agent uses the board's production history and machine readings to investigate likely causes and produce a sourced report for a line manager. Inconclusive inspections and initial handling require operator review. FOX uses an LLM, an agent harness (NemoClaw by default, with alternatives supported), and NVIDIA OpenShell to isolate execution and enforce access policies. Developers connect MES, ERP, machine databases and procedures; fully on-premises operation requires all models and service endpoints to stay within the factory network, while external models require assessment of data sent to them. Physical robot and AMR handling needs factory-specific integrations; the example demonstrates transport in Isaac Sim and ROS 2. Separately, Siemens describes rush-order simulations through Intelligence Center X and virtual AOI preparation using Process Preparation X, Cosmos and Omniverse, but

### Source excerpt

A Siemens report estimates that unplanned downtime costs an average large plant in the surveyed sectors $253 million per year. Whether investigating a failed...

## OpenShell on Kubernetes: A guide for platform engineers

DevFeed: [OpenShell on Kubernetes: A guide for platform engineers](<https://devfeed.tech/articles/openshell-on-kubernetes-a-guide-for-platform-engineers-117173.md>)

Original publisher: [Read original article](<https://www.cncf.io/blog/2026/11/04/openshell-on-kubernetes-a-guide-for-platform-engineers/>)

Author: Jim Bugwadia, Nirmata

Published: 2026-11-04T12:00:00Z

Content type: article

Language: en

Sources: [Cloud Native Computing Foundation](<https://devfeed.tech/sources/cloud-native-computing-foundation.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Kubernetes](<https://devfeed.tech/topics/kubernetes.md>), [Helm charts](<https://devfeed.tech/topics/helm-charts.md>), [container images](<https://devfeed.tech/topics/container-images.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [blog](<https://devfeed.tech/tags/blog.md>), [cncf](<https://devfeed.tech/tags/cncf.md>), [engineers](<https://devfeed.tech/tags/engineers.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [platform](<https://devfeed.tech/tags/platform.md>)

### AI overview

OpenShell runs each agent in a sandbox and routes its outbound requests through a supervisor that can enforce per-binary, per-host, port, and HTTP-method rules, mediate credentials, and accept policy updates without restarting. On Kubernetes it uses SIG Apps' Agent Sandbox, so sandboxes appear as familiar Kubernetes resources; creation currently goes through the Gateway API, and a Supervisor Pod per agent may limit scale. The runtime checks key Pod security settings, while platform teams remain responsible for image, resource, credential, and access controls. The guide pairs RBAC with Kyverno admission policies for those settings and notes that the demo's gateway service account can still create Pods and NetworkPolicies directly, which production policies should cover. Compared with OpenShell's per-request controls and Kubernetes-visible objects, Agent Substrate uses a shared egress gateway, its own API and scheduler, and gVisor or Cloud Hypervisor isolation; it separates agent count from worker-pool Pod count, which may suit many bursty, idle agents. A proposed Substrate isolation backend for OpenShell remains incomplete in the excerpt.

### Source excerpt

What OpenShell is AI agents do more than call APIs. They run shells, install packages, read files, and open network connections. Often they act on text they pulled from the internet a minute earlier. So every...

## "Shall We Approve?" Consequence-Aware Approval for Autonomous Coding Agents

DevFeed: ["Shall We Approve?" Consequence-Aware Approval for Autonomous Coding Agents](<https://devfeed.tech/articles/shall-we-approve-consequence-aware-approval-for-autonomous-coding-agents-111200.md>)

Original publisher: [Read original article](<https://arxiv.org/abs/2610.40753>)

Author: Weiheng Wu, Junkai Chen, Zhensu Sun, Jieke Shi, Yunbo Lyu, David Lo

Published: 2026-11-02T05:00:00Z

Content type: article

Language: en

Sources: [arXiv: Artificial Intelligence](<https://devfeed.tech/sources/arxiv-artificial-intelligence.md>), [arXiv: Cryptography and Security](<https://devfeed.tech/sources/arxiv-cryptography-and-security.md>)

Topics: [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [agentic workflows](<https://devfeed.tech/topics/agentic-workflows.md>), [Command-line interface](<https://devfeed.tech/topics/cli.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>)

Tags: [abstract](<https://devfeed.tech/tags/abstract.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agents](<https://devfeed.tech/tags/agents.md>), [allowed](<https://devfeed.tech/tags/allowed.md>), [arxiv](<https://devfeed.tech/tags/arxiv.md>), [assessment](<https://devfeed.tech/tags/assessment.md>), [automated](<https://devfeed.tech/tags/automated.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [autonomous-coding](<https://devfeed.tech/tags/autonomous-coding.md>), [base](<https://devfeed.tech/tags/base.md>), [based](<https://devfeed.tech/tags/based.md>), [behavior](<https://devfeed.tech/tags/behavior.md>), [benchmark](<https://devfeed.tech/tags/benchmark.md>), [blocking](<https://devfeed.tech/tags/blocking.md>), [build](<https://devfeed.tech/tags/build.md>), [can](<https://devfeed.tech/tags/can.md>), [changes](<https://devfeed.tech/tags/changes.md>), [classifiers](<https://devfeed.tech/tags/classifiers.md>), [coding](<https://devfeed.tech/tags/coding.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [cs-ai](<https://devfeed.tech/tags/cs-ai.md>), [cs-cr](<https://devfeed.tech/tags/cs-cr.md>)

### AI overview

Caper evaluates a shell command's observed effects before deciding whether to allow it: it runs the command in a snapshotted sandbox, gives an independent approver a bounded effects report, then commits allowed commands or rolls back blocked ones with an actionable reason. The abstract reports the lowest harmful-outcome rates among six compared systems on SABER (6.7% versus at least 14.8%) and YoloFS-Ext (7.8% versus at least 35.3%), where harmful behavior is hidden in invoked tools. On YoloFS-Ext, it blocked 1.0% of safe twins of dangerous tasks. These findings are limited to the described benchmarks.

### Source excerpt

arXiv:2610.40753v1 Announce Type: new Abstract: Coding agents increasingly rely on automated approvers to decide whether a proposed shell command may run. Existing approvers, from rule-based checkers to guardrail models and deployed permission classifiers, base their decisions on the text of the request and command. Yet a command's behavior depends on the code it invokes and the state of the workspace, and systems that observe file changes leave the assessment to the agent or a human after execution. We present Caper, a consequence-aware approval mechanism that pre-executes each command in a snapshotted sandbox, provides an independent approver with a bounded report of its observed effects, commits allowed commands, and rolls back blocked ones with a reason the agent can act on. Compared with six existing systems, Caper achieves the lowest harmful-outcome rate on SABER (6.7% versus at least 14.8%) and YoloFS-Ext, a benchmark we build in which harmful behavior is hidden inside invoked tools (7.8% versus at least 35.3%), while blocking 1.0% of the safe twins of the dangerous tasks in YoloFS-Ext.

## Build Specialized Agents on a Stack You Control with NVIDIA NemoClaw

DevFeed: [Build Specialized Agents on a Stack You Control with NVIDIA NemoClaw](<https://devfeed.tech/articles/build-specialized-agents-on-a-stack-you-control-with-nvidia-nemoclaw-109507.md>)

Original publisher: [Read original article](<https://developer.nvidia.com/blog/build-specialized-agents-on-a-stack-you-control-with-nvidia-nemoclaw/>)

Author: Nicola Sessions

Published: 2026-10-30T16:00:00Z

Content type: article

Language: en

Sources: [NVIDIA Developer](<https://devfeed.tech/sources/nvidia-developer.md>), [NVIDIA Technical Blog](<https://devfeed.tech/sources/nvidia-technical-blog.md>)

Topics: [NemoClaw](<https://devfeed.tech/topics/nemoclaw.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Nemotron](<https://devfeed.tech/topics/nemotron.md>), [hermes-agent](<https://devfeed.tech/topics/hermes-agent.md>), [Self Hosted AI Tools](<https://devfeed.tech/topics/self-hosted-ai-tools.md>), [NeMo](<https://devfeed.tech/topics/nemo.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [agent orchestration](<https://devfeed.tech/topics/agent-orchestration.md>), [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [DGX Spark](<https://devfeed.tech/topics/dgx-spark.md>)

Tags: [2](<https://devfeed.tech/tags/2.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agentic-ai-generative-ai](<https://devfeed.tech/tags/agentic-ai-generative-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [build](<https://devfeed.tech/tags/build.md>), [build-ai-agents](<https://devfeed.tech/tags/build-ai-agents.md>), [building](<https://devfeed.tech/tags/building.md>), [control](<https://devfeed.tech/tags/control.md>), [deploy](<https://devfeed.tech/tags/deploy.md>), [developer-tools-techniques](<https://devfeed.tech/tags/developer-tools-techniques.md>), [developers](<https://devfeed.tech/tags/developers.md>), [domain](<https://devfeed.tech/tags/domain.md>), [harness](<https://devfeed.tech/tags/harness.md>), [llms](<https://devfeed.tech/tags/llms.md>), [model](<https://devfeed.tech/tags/model.md>), [need](<https://devfeed.tech/tags/need.md>), [nemoclaw](<https://devfeed.tech/tags/nemoclaw.md>), [nemotron](<https://devfeed.tech/tags/nemotron.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [requires](<https://devfeed.tech/tags/requires.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [stack](<https://devfeed.tech/tags/stack.md>), [tools](<https://devfeed.tech/tags/tools.md>), [with](<https://devfeed.tech/tags/with.md>)

### AI overview

NemoClaw is a set of blueprints that assembles an open agent stack around a developer's chosen harness; it is neither a harness nor a runtime. Its blueprints combine a harness tuned for Nemotron, Nemotron models, NeMo optimization and observability tools, and OpenShell runtime policies governing access to files, networks, credentials, and tools. The text names OpenClaw, Hermes Agent, and LangChain Deep Agents, each with different capabilities, and says existing agents can use a blueprint without being rewritten. As one example, tuning LangChain Deep Agents for Nemotron 3 Ultra reportedly achieved near-frontier performance at about 10x lower cost per run than leading closed models, without retraining. A Hermes Chief of Staff recipe illustrates persistent memory, OpenShell controls, and NeMo Relay traces for audit and optimization. The stack is described as deployable across cloud, data centers, and supported NVIDIA systems; tools and policies can be customized for each workflow.

### Source excerpt

Building an autonomous agent requires a model, a harness, and domain-specific tools. To deploy an agent more securely, developers need a runtime that enforces...

## Red Hat brings OpenShell security to enterprise AI agents

DevFeed: [Red Hat brings OpenShell security to enterprise AI agents](<https://devfeed.tech/articles/red-hat-brings-openshell-security-to-enterprise-ai-agents-102765.md>)

Original publisher: [Read original article](<https://siliconangle.com/2026/10/27/red-hat-openclaw-agents-kubeconna/>)

Author: Devony Hof

Published: 2026-10-28T00:09:09Z

Content type: news

Language: en

Sources: [SiliconANGLE](<https://devfeed.tech/sources/siliconangle.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [ai security](<https://devfeed.tech/topics/ai-security.md>), [redhat](<https://devfeed.tech/topics/redhat.md>), [Access Control](<https://devfeed.tech/topics/access-control.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-sandbox](<https://devfeed.tech/tags/agent-sandbox.md>), [agent-security](<https://devfeed.tech/tags/agent-security.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-inference](<https://devfeed.tech/tags/ai-inference.md>), [ai-infrastructure](<https://devfeed.tech/tags/ai-infrastructure.md>), [ai-platform](<https://devfeed.tech/tags/ai-platform.md>), [ai-powered](<https://devfeed.tech/tags/ai-powered.md>), [ai-security](<https://devfeed.tech/tags/ai-security.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [cloud](<https://devfeed.tech/tags/cloud.md>), [cloud-native](<https://devfeed.tech/tags/cloud-native.md>), [cloud-native-computing-foundation](<https://devfeed.tech/tags/cloud-native-computing-foundation.md>), [cloud-security](<https://devfeed.tech/tags/cloud-security.md>), [cloudnativecon](<https://devfeed.tech/tags/cloudnativecon.md>), [cncf](<https://devfeed.tech/tags/cncf.md>), [cube-event-coverage](<https://devfeed.tech/tags/cube-event-coverage.md>), [deployments](<https://devfeed.tech/tags/deployments.md>), [developer-platforms](<https://devfeed.tech/tags/developer-platforms.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [enterprise-ai](<https://devfeed.tech/tags/enterprise-ai.md>), [gpu-infrastructure](<https://devfeed.tech/tags/gpu-infrastructure.md>), [kubecon](<https://devfeed.tech/tags/kubecon.md>), [kubecon-plus-cloudnativecon-na-2026](<https://devfeed.tech/tags/kubecon-plus-cloudnativecon-na-2026.md>), [kubeconna](<https://devfeed.tech/tags/kubeconna.md>), [kubeconna26eventpage](<https://devfeed.tech/tags/kubeconna26eventpage.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [news](<https://devfeed.tech/tags/news.md>), [nyse-ibm](<https://devfeed.tech/tags/nyse-ibm.md>), [observability](<https://devfeed.tech/tags/observability.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [openclaw](<https://devfeed.tech/tags/openclaw.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [pete-cheslock](<https://devfeed.tech/tags/pete-cheslock.md>), [platform-engineering](<https://devfeed.tech/tags/platform-engineering.md>), [red-hat](<https://devfeed.tech/tags/red-hat.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [security](<https://devfeed.tech/tags/security.md>), [stu-miniman](<https://devfeed.tech/tags/stu-miniman.md>), [thecube](<https://devfeed.tech/tags/thecube.md>)

### AI overview

Red Hat is integrating Nvidia-developed OpenShell into its AI platform to help secure and govern OpenClaw agents in enterprise settings. OpenShell places agents in isolated sandboxes and monitors access to potentially vulnerable information. Its agent sandbox builds on Kubernetes-native APIs, allowing engineers to deploy agents without learning a separate set of deployment APIs, though the excerpt notes they still have new concepts to learn. The article also describes AI security tools as a way to test for vulnerabilities at greater speed and scale; it does not detail specific controls or compare products.

### Source excerpt

After OpenClaw made a splash -- or a tidal wave, depending on who you ask -- in the artificial intelligence community, security became paramount. Ahead of the upcoming KubeCon + CloudNativeCon NA event, Red Hat LLC leaders gave more insight into OpenShell, a secure, open runtime for agents that is designed to keep OpenClaw deployments safe [...] The post Red Hat brings OpenShell security to enterprise AI agents appeared first on SiliconANGLE.

## Three layers of open source defense for AI coding agents: OpenShell, Prempti, and Falco

DevFeed: [Three layers of open source defense for AI coding agents: OpenShell, Prempti, and Falco](<https://devfeed.tech/articles/three-layers-of-open-source-defense-for-ai-coding-agents-openshell-prempti-and-falco-102877.md>)

Original publisher: [Read original article](<https://webflow.sysdig.com/blog/three-layers-of-open-source-defense-for-ai-coding-agents-openshell-prempti-and-falco>)

Author: Álvaro Iradier

Published: 2026-10-27T00:00:00Z

Content type: article

Language: en

Sources: [Sysdig](<https://devfeed.tech/sources/sysdig-blog.md>)

Topics: [Falco](<https://devfeed.tech/topics/falco.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [agentic-coding](<https://devfeed.tech/topics/agentic-coding.md>), [Open Source](<https://devfeed.tech/topics/open-source.md>)

Tags: [ai](<https://devfeed.tech/tags/ai.md>), [coding-agents](<https://devfeed.tech/tags/coding-agents.md>), [defense](<https://devfeed.tech/tags/defense.md>), [falco](<https://devfeed.tech/tags/falco.md>), [layers](<https://devfeed.tech/tags/layers.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [prempti](<https://devfeed.tech/tags/prempti.md>)

### AI overview

The demo shows how three open source layers cover different risks from AI coding agents. NVIDIA OpenShell enforces filesystem and network allowlists outside the agent; Prempti judges Claude Code tool calls and can allow, ask, or deny; Falco observes host syscalls, including failed attempts and tampering. Tests show Prempti can block an explicit credential check, while an indirect Python script passes its tool-call checks and OpenShell policy but triggers Falco when it opens the credential path. OpenShell also blocks direct network access and sandbox tampering, while Falco records those attempts. Prempti's in-sandbox daemon can be stopped or hijacked by the agent; moving its hook to root-owned managed settings and asking before signals makes daemon failure fail closed, but does not prevent socket replacement. The authors report about 160 alerts in an ordinary Claude Code session, reduced to zero with environment changes and narrow rule exceptions, underscoring the need to tune detection against real sessions.

### Source excerpt

Three open source defense layers for AI coding agents: OpenShell sandboxes, Prempti judges tool calls, Falco watches syscalls. See what each catches and misses.

## Three layers, all config: how DigitalOcean Managed Agents enforce NVIDIA OpenShell-aligned boundaries

DevFeed: [Three layers, all config: how DigitalOcean Managed Agents enforce NVIDIA OpenShell-aligned boundaries](<https://devfeed.tech/articles/three-layers-all-config-how-digitalocean-managed-agents-enforce-nvidia-openshell-aligned-boundaries-100676.md>)

Original publisher: [Read original article](<https://www.digitalocean.com/blog/managed-agents-enforce-nvidia-openshell>)

Author: Rohit Bhattacharjee

Published: 2026-10-26T17:21:41Z

Content type: article

Language: en

Sources: [DigitalOcean](<https://devfeed.tech/sources/digitalocean.md>)

Topics: [Digital Ocean](<https://devfeed.tech/topics/digital-ocean.md>), [Agent Harness](<https://devfeed.tech/topics/agent-harness.md>), [AI Infrastructure](<https://devfeed.tech/topics/ai-infrastructure.md>), [human in the loop agent](<https://devfeed.tech/topics/human-in-the-loop-agent.md>), [Supabase](<https://devfeed.tech/topics/supabase.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>)

Tags: [agents](<https://devfeed.tech/tags/agents.md>), [agents-in-production](<https://devfeed.tech/tags/agents-in-production.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-native](<https://devfeed.tech/tags/ai-native.md>), [config](<https://devfeed.tech/tags/config.md>), [data](<https://devfeed.tech/tags/data.md>), [database](<https://devfeed.tech/tags/database.md>), [digitalocean](<https://devfeed.tech/tags/digitalocean.md>), [digitalocean-managed-agents](<https://devfeed.tech/tags/digitalocean-managed-agents.md>), [embeddings](<https://devfeed.tech/tags/embeddings.md>), [firecracker](<https://devfeed.tech/tags/firecracker.md>), [gateway](<https://devfeed.tech/tags/gateway.md>), [infrastructure](<https://devfeed.tech/tags/infrastructure.md>), [managed](<https://devfeed.tech/tags/managed.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [nvidia-openshell](<https://devfeed.tech/tags/nvidia-openshell.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [partner-news](<https://devfeed.tech/tags/partner-news.md>)

### AI overview

A DigitalOcean Managed Agent migrated huffl's production backend from Supabase, handling a database with 31 tables and 145,432 rows, three storage buckets, and about 31,000 vector embeddings. The excerpt says the agent provisioned the database, loaded access policies, mapped Supabase features to DigitalOcean equivalents, and verified data table by table. DigitalOcean Managed Agents is in public preview, without an SLA and subject to preview terms. Its Harness Runtime gives each session a Firecracker microVM that can pause, resume, and fork; Action Gateway provides a managed MCP endpoint to more than 16,000 tools, brokers credentials at call time, and offers human approval for sensitive actions. The platform is described as open by default, letting users bring their own harness, model, or key. The excerpt cuts off before describing the three boundary layers in detail.

### Source excerpt

A few weeks ago, we handed an AI agent one of the scariest jobs in infrastructure: moving huffl's production backend from Supabase to DigitalOcean. huffl.ai is a real product. Underneath it sits 31 tables, 145,432 rows, three storage buckets, and about 31,000 vector embeddings. A DigitalOcean Managed Agent did the migration, directed from a phone in plain chat. It provisioned the database, loaded the access policies, mapped every Supabase feature to its DigitalOcean counterpart, and verified the data table by table. There is something magical about watching that happen. A message typed while waiting for coffee. A decision made between meetings. No laptop, no terminal, no maintenance window. By the end of the week the backend had moved, and nobody worked a late night. But magic is not a security model. For days, the agent had the run of two clouds and touched production data the entire time. What is worth examining is how it stayed inside its boundaries while doing it: how to hand an agent that much power without handing it a key, and how to draw lines it cannot cross. Three layers. All config. What is DigitalOcean Managed Agents? DigitalOcean Managed Agents is our platform for running AI agents in production, now in public preview, so you don't have to build the plumbing yourself. Public preview products are provided without a service level agreement and are subject to the DigitalOcean preview terms; features may change or be withdrawn before general availability. DigitalOcean Harness Runtime gives every session its own Firecracker microVM, and sessions keep running, pause, resume, and fork even after you close your laptop. Action Gateway gives agents one managed MCP endpoint to more than 16,000 tools as of October 2026, with credentials brokered at the moment of the call and human approval available for sensitive actions. It runs on the DigitalOcean AI-Native Cloud and is open by default, so you bring your own harness, model, or key. Here's how it stayed inside its

## Is Kubernetes ready for AI agents?

DevFeed: [Is Kubernetes ready for AI agents?](<https://devfeed.tech/articles/is-kubernetes-ready-for-ai-agents-99192.md>)

Original publisher: [Read original article](<https://thenewstack.io/kubernetes-ready-ai-agents/>)

Author: Dan Ciruli

Published: 2026-10-25T14:00:00Z

Content type: article

Language: en

Sources: [Kubernetes Overview, News and Trends | The New Stack](<https://devfeed.tech/sources/kubernetes-overview-news-and-trends-the-new-stack.md>), [The New Stack](<https://devfeed.tech/sources/the-new-stack.md>)

Topics: [Amazon Elastic Kubernetes Service](<https://devfeed.tech/topics/amazon-elastic-kubernetes-service.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Amazon Bedrock AgentCore](<https://devfeed.tech/topics/amazon-bedrock-agentcore.md>)

Tags: [abstraction](<https://devfeed.tech/tags/abstraction.md>), [agent](<https://devfeed.tech/tags/agent.md>), [agentic](<https://devfeed.tech/tags/agentic.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agentic-platforms](<https://devfeed.tech/tags/agentic-platforms.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-inference](<https://devfeed.tech/tags/ai-inference.md>), [ai-infrastructure](<https://devfeed.tech/tags/ai-infrastructure.md>), [ai-labs](<https://devfeed.tech/tags/ai-labs.md>), [ai-models](<https://devfeed.tech/tags/ai-models.md>), [ai-platform](<https://devfeed.tech/tags/ai-platform.md>), [ai-sandboxing](<https://devfeed.tech/tags/ai-sandboxing.md>), [ai-workloads](<https://devfeed.tech/tags/ai-workloads.md>), [application](<https://devfeed.tech/tags/application.md>), [applications](<https://devfeed.tech/tags/applications.md>), [cios](<https://devfeed.tech/tags/cios.md>), [cncf](<https://devfeed.tech/tags/cncf.md>), [developers](<https://devfeed.tech/tags/developers.md>), [kubernetes](<https://devfeed.tech/tags/kubernetes.md>), [post-contributed](<https://devfeed.tech/tags/post-contributed.md>), [software](<https://devfeed.tech/tags/software.md>), [sponsor-cncf](<https://devfeed.tech/tags/sponsor-cncf.md>), [sponsored-post-contributed](<https://devfeed.tech/tags/sponsored-post-contributed.md>)

### AI overview

Kubernetes can schedule, scale, and recover AI inference workloads, and the excerpt says two-thirds of organizations running inference use it. But agent behavior creates security and runtime demands beyond conventional services: the article calls for isolated sandboxes, explicit permissions for files, tools, network and system resources, and controls on code execution. It says CNCF work is underway, while open-source Kubernetes does not yet provide the full isolation needed, which may require virtual machines or host-level controls; some commercial distributions already offer such capabilities. Agents may run for seconds and need to start quickly, making fresh pod startup too slow for interactive use. An "agent substrate" warm pool is presented as one approach, keeping agents suspended so they can resume in milliseconds. The excerpt also identifies monitoring as an enterprise concern.

### Source excerpt

Kubernetes was built around a fairly simple assumption. It's that software does what its developers told it to do. AI The post Is Kubernetes ready for AI agents? appeared first on The New Stack.

## Meet Anthropic's sandbox-runtime: OS-Level Sandboxing for AI Agents, No Containers

DevFeed: [Meet Anthropic's sandbox-runtime: OS-Level Sandboxing for AI Agents, No Containers](<https://devfeed.tech/articles/meet-anthropic-s-sandbox-runtime-os-level-sandboxing-for-ai-agents-no-containers-98776.md>)

Original publisher: [Read original article](<https://www.marktechpost.com/2026/10/24/anthropic-sandbox-runtime-ai-agent-sandbox/>)

Author: Asif Razzaq

Published: 2026-10-24T11:12:27Z

Content type: article

Language: en

Sources: [MarkTechPost](<https://devfeed.tech/sources/marktechpost.md>)

Topics: [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Claude](<https://devfeed.tech/topics/claude.md>), [Node.js](<https://devfeed.tech/topics/node-js.md>), [Ubuntu](<https://devfeed.tech/topics/ubuntu.md>)

Tags: [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-shorts](<https://devfeed.tech/tags/ai-shorts.md>), [anthropic](<https://devfeed.tech/tags/anthropic.md>), [applications](<https://devfeed.tech/tags/applications.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [claude](<https://devfeed.tech/tags/claude.md>), [claude-code](<https://devfeed.tech/tags/claude-code.md>), [cli](<https://devfeed.tech/tags/cli.md>), [containers](<https://devfeed.tech/tags/containers.md>), [context-engineering](<https://devfeed.tech/tags/context-engineering.md>), [editors-pick](<https://devfeed.tech/tags/editors-pick.md>), [file](<https://devfeed.tech/tags/file.md>), [for-devs](<https://devfeed.tech/tags/for-devs.md>), [harness](<https://devfeed.tech/tags/harness.md>), [linux](<https://devfeed.tech/tags/linux.md>), [macos](<https://devfeed.tech/tags/macos.md>), [mcp](<https://devfeed.tech/tags/mcp.md>), [mcp-servers](<https://devfeed.tech/tags/mcp-servers.md>), [network](<https://devfeed.tech/tags/network.md>), [new-releases](<https://devfeed.tech/tags/new-releases.md>), [open-source](<https://devfeed.tech/tags/open-source.md>), [os](<https://devfeed.tech/tags/os.md>), [runtime](<https://devfeed.tech/tags/runtime.md>), [sandbox](<https://devfeed.tech/tags/sandbox.md>), [sandboxing](<https://devfeed.tech/tags/sandboxing.md>), [seatbelt](<https://devfeed.tech/tags/seatbelt.md>), [software-engineering](<https://devfeed.tech/tags/software-engineering.md>), [staff](<https://devfeed.tech/tags/staff.md>), [tech-news](<https://devfeed.tech/tags/tech-news.md>), [technology](<https://devfeed.tech/tags/technology.md>)

### AI overview

Anthropic's open-source sandbox-runtime (srt) wraps processes in OS-level file and network restrictions, including their subprocesses, without containers. It can run as a CLI or TypeScript library for agents, MCP servers, and other commands. macOS uses Seatbelt; Linux uses bubblewrap and requires socat and ripgrep; Windows support is alpha and uses a dedicated account, file ACLs, and network filtering. Reads are allowed by default unless denied, writes are denied unless allowed, and network access requires domain allowlisting through host-side proxies. Settings support wildcards, per-command allowances, prompts for unmatched hosts, live rule updates, and checks against loopback or cloud metadata addresses. The practical tradeoff is that restrictive allowlists can block tasks such as dependency installation, while broad rules weaken protection. Anthropic reports 84% fewer Claude Code permission prompts in internal usage; this is self-reported, not an external benchmark. The excerpt describes srt but provides no actual comparison with other sandboxes.

### Source excerpt

Anthropic sandbox-runtime (srt) is the open-source, OS-level sandbox behind Claude Code. It restricts file and network access for agents, MCP servers and CLI tools with Seatbelt on macOS and bubblewrap on Linux, no containers needed. The post Meet Anthropic's sandbox-runtime: OS-Level Sandboxing for AI Agents, No Containers appeared first on MarkTechPost.

## Nvidia's Berlin GTC: Faster AI factories, lower power use and more secure agents

DevFeed: [Nvidia's Berlin GTC: Faster AI factories, lower power use and more secure agents](<https://devfeed.tech/articles/nvidia-s-berlin-gtc-faster-ai-factories-lower-power-use-and-more-secure-agents-96168.md>)

Original publisher: [Read original article](<https://www.networkworld.com/article/4239248/nvidias-berlin-gtc-faster-ai-factories-lower-power-use-and-more-secure-agents.html>)

Author: Andy Patrizio

Published: 2026-10-22T17:51:50Z

Content type: news

Language: en

Sources: [Network World](<https://devfeed.tech/sources/network-world.md>)

Topics: [NVIDIA DSX](<https://devfeed.tech/topics/nvidia-dsx.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Isaac](<https://devfeed.tech/topics/isaac.md>), [clean energy data centers](<https://devfeed.tech/topics/clean-energy-data-centers.md>)

Tags: [accelerate](<https://devfeed.tech/tags/accelerate.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [ai-compute](<https://devfeed.tech/tags/ai-compute.md>), [ai-factories](<https://devfeed.tech/tags/ai-factories.md>), [ai-factory](<https://devfeed.tech/tags/ai-factory.md>), [ai-infrastructure](<https://devfeed.tech/tags/ai-infrastructure.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [artificial-intelligence-data-center-robotics](<https://devfeed.tech/tags/artificial-intelligence-data-center-robotics.md>), [autonomous](<https://devfeed.tech/tags/autonomous.md>), [autonomous-ai-agents](<https://devfeed.tech/tags/autonomous-ai-agents.md>), [berlin](<https://devfeed.tech/tags/berlin.md>), [bluefield-4](<https://devfeed.tech/tags/bluefield-4.md>), [compute](<https://devfeed.tech/tags/compute.md>), [conference](<https://devfeed.tech/tags/conference.md>), [cooling](<https://devfeed.tech/tags/cooling.md>), [data-center](<https://devfeed.tech/tags/data-center.md>), [edition-us](<https://devfeed.tech/tags/edition-us.md>), [faster](<https://devfeed.tech/tags/faster.md>), [gtc](<https://devfeed.tech/tags/gtc.md>), [language-en](<https://devfeed.tech/tags/language-en.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [power](<https://devfeed.tech/tags/power.md>), [robotics](<https://devfeed.tech/tags/robotics.md>), [storytype-news](<https://devfeed.tech/tags/storytype-news.md>)

### AI overview

At its Berlin GTC, Nvidia announced a modular data center system intended to shorten AI infrastructure deployment from months to weeks. Its DSX reference architecture combines liquid cooling, dry coolers, and power-management software; Nvidia claims DSX MaxLPS can produce the same number of AI tokens with up to 30% less power, and estimates dry cooling at 2 gigawatts could save 54 million liters of water daily versus cooling towers. OpenShell combines Vera Rubin compute, Vera CPUs, BlueField-4 DPUs, and security software to build, monitor, and constrain agents under security policies; the reference architecture is available now. Nvidia also expanded Isaac robotics tools, while Agile Robots' Jetson Thor-powered humanoid platform is planned for early 2027. In telecom, T-Mobile completed what the companies called the first AI-RAN call on a commercial network, and BT, Deutsche Telekom, and Elisa announced pilots. Efficiency figures and deployment benefits are vendor claims.

### Source excerpt

Nvidia is hosting its GTC conference in Berlin this week, and while the bulk of the news is European-oriented, there are several stories of interest around the world. They include the usual stories of Nvidia: AI compute platforms, data center infrastructure, and networking. Nvidia Unveils DSX POD Modular AI Factories Nvidia introduced the DSX POD, a factory-built, modular data center approach designed to accelerate AI infrastructure deployment. The concept packages compute and supporting infrastructure into integrated systems rather than requiring every deployment to be built from scratch. Nvidia is really positioning this as a turnkey solution, stating that it could reduce deployment timelines from months to weeks. "An AI factory creates value when it starts producing tokens -- not when construction is complete," said Jensen Huang, founder and CEO of Nvidia in a statement. "With Nvidia DSX POD, we're bringing factory-built precision and speed to AI infrastructure construction -- reducing deployment time from months to weeks so customers can scale capacity and bring revenue-generating production online faster." Nvidia Targets Lower Power Consumption and Water Use Nvidia's DSX reference architecture combines closed-loop liquid cooling, dry coolers and power-management software. The company says its DSX MaxLPS software can generate the same number of AI tokens using up to 30% less power. Nvidia also estimates that a 2-gigawatt deployment using its dry-cooling approach could save 54 million liters of cooling water per day compared with conventional cooling-tower systems. OpenShell Brings Security Controls to Autonomous AI Agents Nvidia announced its OpenShell Reference Architecture, combining Vera Rubin compute, Vera CPUs, BlueField-4 data processing units (DPUs) and security software to build, monitor and constrain AI agents. The OpenShell Reference Architecture builds on the foundation of Nvidia OpenShell secure runtime software, extending it into full-stack, industria

## VAST Agent Sandbox Foundation Stack Puts NVIDIA OpenShell Sandboxes on Zero-Copy Clones and Logs Every Tool Call

DevFeed: [VAST Agent Sandbox Foundation Stack Puts NVIDIA OpenShell Sandboxes on Zero-Copy Clones and Logs Every Tool Call](<https://devfeed.tech/articles/vast-agent-sandbox-foundation-stack-puts-nvidia-openshell-sandboxes-on-zero-copy-clones-and-logs-every-tool-call-95846.md>)

Original publisher: [Read original article](<https://www.storagereview.com/news/vast-agent-sandbox-foundation-stack-nvidia-openshell-nemoclaw-zero-copy-clones>)

Author: Harold Fritts

Published: 2026-10-22T14:16:10Z

Content type: news

Language: en

Sources: [StorageReview.com](<https://devfeed.tech/sources/storagereview-com.md>)

Topics: [NemoClaw](<https://devfeed.tech/topics/nemoclaw.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Multi Agent Systems](<https://devfeed.tech/topics/multi-agent-systems.md>), [AI Context Management](<https://devfeed.tech/topics/ai-context-management.md>), [long-running](<https://devfeed.tech/topics/long-running.md>), [supervisor agent pattern](<https://devfeed.tech/topics/supervisor-agent-pattern.md>), [telemetry](<https://devfeed.tech/topics/telemetry.md>), [NVIDIA NIM](<https://devfeed.tech/topics/nvidia-nim.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-fleets](<https://devfeed.tech/tags/agent-fleets.md>), [agent-memory](<https://devfeed.tech/tags/agent-memory.md>), [agent-sandbox](<https://devfeed.tech/tags/agent-sandbox.md>), [agents](<https://devfeed.tech/tags/agents.md>), [agents-in-production](<https://devfeed.tech/tags/agents-in-production.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agents](<https://devfeed.tech/tags/ai-agents.md>), [audit](<https://devfeed.tech/tags/audit.md>), [audit-trail](<https://devfeed.tech/tags/audit-trail.md>), [clone](<https://devfeed.tech/tags/clone.md>), [enterprise](<https://devfeed.tech/tags/enterprise.md>), [execution](<https://devfeed.tech/tags/execution.md>), [nemoclaw](<https://devfeed.tech/tags/nemoclaw.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [nvidia-openshell](<https://devfeed.tech/tags/nvidia-openshell.md>), [openshell](<https://devfeed.tech/tags/openshell.md>), [policy-enforcement](<https://devfeed.tech/tags/policy-enforcement.md>), [reference-architecture](<https://devfeed.tech/tags/reference-architecture.md>), [sandboxes](<https://devfeed.tech/tags/sandboxes.md>), [stack](<https://devfeed.tech/tags/stack.md>), [zero-copy](<https://devfeed.tech/tags/zero-copy.md>)

### AI overview

VAST and NVIDIA's Agent Sandbox Foundation Stack separates agents' durable state and records from their temporary compute. NVIDIA OpenShell provides isolated execution, identity-based access policies, and logging of interactions outside each sandbox; VAST's DataStore Block CSI driver provisions persistent volumes from zero-copy clones, while snapshots support suspend and resume. VAST DataBase serves as shared vector storage and retains agent memory, traces, reports, and records of tool calls, prompts, tokens, costs, and decisions. In the published Apache 2.0 reference implementation, an MCP server is the sandbox's only route to the database, and sub-agents receive clones of a workflow's head volume. The stack also combines VAST fleet telemetry with NVIDIA Sentry; VAST says flagged agents can have data access isolated and sandbox state frozen for review. The implementation demonstrates a supervised research workflow with parallel sandboxed sub-agents. The article says the design targets thousands of agents, but provides no measured provisioning or concurrency results, general availability date, software versions, or hardware configurations.

### Source excerpt

VAST Data announced the Agent Sandbox Foundation Stack with NVIDIA at NVIDIA GTC Berlin, a reference architecture that runs NVIDIA NemoClaw agents inside NVIDIA OpenShell sandboxes on the VAST AI Operating System. OpenShell handles isolated execution and policy enforcement around each agent, and VAST handles the parts that have to outlive the sandbox: the agent's The post VAST Agent Sandbox Foundation Stack Puts NVIDIA OpenShell Sandboxes on Zero-Copy Clones and Logs Every Tool Call appeared first on StorageReview.com.

## Nvidia turns the AI factory into a product, and agents into the workload

DevFeed: [Nvidia turns the AI factory into a product, and agents into the workload](<https://devfeed.tech/articles/nvidia-turns-the-ai-factory-into-a-product-and-agents-into-the-workload-93180.md>)

Original publisher: [Read original article](<https://siliconangle.com/2026/10/21/nvidia-turns-the-ai-factory-into-a-product-and-agents-into-the-workload/>)

Author: Zeus Kerravala

Published: 2026-10-21T20:52:15Z

Content type: news

Language: en

Sources: [SiliconANGLE](<https://devfeed.tech/sources/siliconangle.md>)

Topics: [AI Factory](<https://devfeed.tech/topics/ai-factory.md>), [NVIDIA DSX](<https://devfeed.tech/topics/nvidia-dsx.md>), [OpenShell](<https://devfeed.tech/topics/openshell.md>), [Nvidia](<https://devfeed.tech/topics/nvidia.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [agent-security](<https://devfeed.tech/tags/agent-security.md>), [agentic-ai](<https://devfeed.tech/tags/agentic-ai.md>), [agents](<https://devfeed.tech/tags/agents.md>), [ai](<https://devfeed.tech/tags/ai.md>), [ai-agent](<https://devfeed.tech/tags/ai-agent.md>), [ai-chip](<https://devfeed.tech/tags/ai-chip.md>), [ai-factories](<https://devfeed.tech/tags/ai-factories.md>), [ai-infrastructure](<https://devfeed.tech/tags/ai-infrastructure.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [artificial-intelligence](<https://devfeed.tech/tags/artificial-intelligence.md>), [bluefield-4](<https://devfeed.tech/tags/bluefield-4.md>), [chip](<https://devfeed.tech/tags/chip.md>), [chips](<https://devfeed.tech/tags/chips.md>), [cisco-systems](<https://devfeed.tech/tags/cisco-systems.md>), [generative-ai](<https://devfeed.tech/tags/generative-ai.md>), [gtc](<https://devfeed.tech/tags/gtc.md>), [infra](<https://devfeed.tech/tags/infra.md>), [jensen-huang](<https://devfeed.tech/tags/jensen-huang.md>), [news](<https://devfeed.tech/tags/news.md>), [nvidia](<https://devfeed.tech/tags/nvidia.md>), [nvidia-gtc](<https://devfeed.tech/tags/nvidia-gtc.md>), [the-latest](<https://devfeed.tech/tags/the-latest.md>), [top-story-5](<https://devfeed.tech/tags/top-story-5.md>)

### AI overview

At GTC Berlin, Nvidia presented a strategy to sell an AI factory operating model alongside its chips, addressing deployment time, power, agent security, enterprise data and telecom infrastructure. Its DSX POD reference design packages seven Vera Rubin NVL72 racks with cabling and cooling; Nvidia says a prepared-site deployment could shrink from six months to as little as 14 days, with delivery expected in early 2027. That estimate assumes site readiness and does not remove grid connection delays. Nvidia also says DSX MaxLPS can provide the same output with up to 30% less power, while DSX Flex can defer or throttle workloads in response to grid needs. OpenShell places policy enforcement outside an agent's reach; the article distinguishes its software protections from additional hardware isolation using BlueField-4 and Sentry. Nvidia's Agent Knowledge Foundation and Kumo structured-data models target enterprise tables, and the company is promoting telco sites as distributed AI infrastructure. The article frames these launches as a broader bet on faster deployment and time to revenue, while noting that Nvidia's promise of an open stack will face scrutiny as its offerings expand.

### Source excerpt

For most of the past half-decade, the narrative around Nvidia Corp. has largely centered on the graphics processing unit. The questions have been who has allocation, how much faster the next generation is and whether anyone can close the gap. At GTC Berlin early today, Chief Executive Jensen Huang spent surprisingly little of his keynote [...] The post Nvidia turns the AI factory into a product, and agents into the workload appeared first on SiliconANGLE.

[Next page](<https://devfeed.tech/topics/openshell.md?cursor=WyIyMDI2LTEwLTIxVDIwOjUyOjE1KzAwOjAwIiwgIjc0MTY1MWFiLTQ5NzMtNGI1ZC05YmNiLWYxZjA2YmNhNGZkNiJd>)