November 10 agent guide: Gemini Robotics 2.5 and OpenShell
November 10, 2026 moved agentics into factories and Kubernetes. Google DeepMind released Gemini Robotics 2.5 for partner robots, and NVIDIA contributed OpenShell to the CNCF as a sandbox project for securing autonomous agents.

In this post (7 sections)
November 10, 2026 is a physical-agent and agent-runtime day for Bacancy Technology readers. Google DeepMind introduced Gemini Robotics 2.5, its most capable vision-language-action model to date for partner robots. At KubeCon, NVIDIA contributed OpenShell to the CNCF Sandbox so organizations can govern autonomous agents with policy outside the model.
What shipped on November 10
| Release | Source | Status | Builder action |
|---|---|---|---|
| Gemini Robotics 2.5 | Google DeepMind | Partner rollout | Track if your robotics OEM is in the partner set; do not assume public API |
| OpenShell → CNCF Sandbox | NVIDIA / CNCF | Open governance under ai-openshell | Evaluate OpenShell or Cedar policies for agent sandboxes |
| Agent Substrate (related) | Google-led / CNCF | Sandbox acceptance noted by NVIDIA | Watch for long-running agent workloads on Kubernetes |
Gemini Robotics 2.5
Gemini Robotics 2.5 is a vision-language-action model aimed at broader skills, faster learning, higher dexterity, and safer physical behavior. DeepMind evaluated more than 200 tasks, from tool use and deformable objects to delicate manipulation. Versus prior baselines, DeepMind reports 30% to 100% higher success rates by task type and 20% to 30% faster completion. The model works across robot types including Apptronik Apollo 2 and Franka Duo platforms.
Factory proof point and safety
- Mercedes-Benz production pilots in Hungary and Germany used Apollo 2 for automotive kitting with a reported 96% success rate.
- Inference-time reasoning guidance steers around physical hazards; DeepMind cites +33% safe completion around obstacles and +43% around nearby people.
- ASIMOV-Jailbreak, a new open-source robotics jailbreak benchmark, is released; DeepMind reports Gemini 4 Argon at 99.5% jailbreak resistance on that suite versus 82% for a prior baseline.
Gemini 4 Argon also appears as a research accelerator: generating simulated practice from camera scenes, building digital twins, and helping diagnose failures. Bacancy Technology reads Robotics 2.5 as embodied agent progress for industrial partners, not a drop-in replacement for software coding agents.
OpenShell joins the CNCF
OpenShell manages and monitors agent behavior and enforces policy-based boundaries on what an agent can reach: files, networks, credentials, and inference endpoints. CNCF and NVIDIA emphasize enforcement in the infrastructure around the agent, not trust in the model alone. OpenShell is adopting Cedar as its policy language. The project moves toward the ai-openshell GitHub organization under CNCF Sandbox governance while NVIDIA continues core engineering.
Builder checklist
- 01Separate physical-agent and software-agent roadmapsDo not force Robotics 2.5 into a software backlog. Ask OEM partners for harness access and safety docs if physical automation is in scope.
- 02Map current agent privilegesList file, network, secret, and model-endpoint access for each coding or browser agent before comparing OpenShell policies.
- 03Watch CNCF working groupsFollow ai-openshell RFCs for observability, multi-tenancy, and Cedar policy patterns rather than forking an internal one-off sandbox.
Conclusion
November 10 connects factory-floor VLAs with Kubernetes-native agent runtimes. DeepMind’s Robotics 2.5 is partner-first physical intelligence. OpenShell’s CNCF move is the governance story software agent teams should track this week.
Related reading
- November 9 Fable 5.2 and Muse Ember: /blog/november-09-claude-fable-5-2-muse-ember-2026
- November 7 OpenCode Browse: /blog/november-07-opencode-browse-browser-agents-2026
- November 1 Mage-VL-Cowork: /blog/november-01-mage-vl-cowork-computer-use-2026
- MCP topic map: /topics/mcp
- Agentic AI topic map: /topics/agentic-ai
Need agent sandbox and policy design?
Bacancy Technology designs agent runtime controls, MCP allowlists, and production sandbox patterns. Consulting: /consulting/ . Contact: /contact/ .
Sources: deepmind.google ; blogs.nvidia.com — open source openshell infrastructure ; cncf.io .
Agentic AI patterns, delivered Thursdays
What I am shipping, watching, and pruning out of client stacks each week. One email. No fluff.