Nebius
Official@nebius · Netherlands
Offers specialized infrastructure and orchestration for Physical AI, robotics simulation, and GPU-accelerated world model training on cloud-native clusters.
Agent Skills by Nebius
Showing 127 vetted skills indexed across 3 GitHub repositories.
nebiussdlc-prepare-execution
Prepares isolated integration worktrees and deterministic dependency waves for locked SDLC feature plans.
nebiuscommit
Creates one validated local Git commit for the complete repository diff without pushing.
nebiusai-agent-design
Designs provider-neutral production AI agent subsystems with contracts, policies, and governance.
nebiusnebius-grafana-query
Run read-only Grafana MCP queries for Nebius metrics, logs, traces, and dashboards.
nebiussdlc-update-documents
Updates project-facing documentation from implemented and evaluated Agentic SDLC evidence.
nebiusterraform
Generate and harden Terraform modules and environment repositories with standards and CI guardrails.
nebiusdesign
Design software features and architectures, then produce an implementation-ready plan handoff.
nebiussdlc-create-design
Convert requirements and context into evidence-backed FEAT design records in docs/design.md.
nebiussdlc-gui-test
Executes browser GUI flows and captures screenshots, accessibility snapshots, and pass/fail evidence against acceptance criteria.
nebiuspython-project
Scaffold and harden Python repositories with uv, pyproject.toml, Ruff, pytest, and CI.
nebiussdlc-uat-tests
Runs product-level acceptance testing across all features before PR creation in the Agentic SDLC workflow.
nebiusoptimize-pytest
Measure and optimize slow pytest suites by ranking cumulative startup, fixture, and teardown costs.
nebiustask-implementer
Coordinates brownfield implementation through dependency waves of isolated Git worktree workers.
nebiussdlc-tdd
Writes feature tests in an integration worktree before implementation within the Agentic SDLC workflow.
nebiusfrontend-project
Generate deterministic React, TypeScript, and Vite project files with validated candidate manifests.
nebiustroubleshoot
Diagnose and repair persistent software and infrastructure failures through causal investigation.
nebiussdlc-evaluate
Evaluate observed product behavior against acceptance criteria and emit normalized failure events.
nebiuscode-info
Generates read-only Markdown metrics reports for local folders or GitHub repositories.
nebiusbrainstorm
Explores ideas in chat using source-ranked context before implementation begins.
nebiusnosleep4mac
Installs and maintains a per-user macOS LaunchAgent running caffeinate -s to prevent system sleep on AC power.
nebiuslinter
Lint and auto-fix Shell, Markdown, and Python files with fallback config rules.
nebiusapp-stack
Selects the smallest justified application stack and coordinates specialist skills for implementation.
nebiusagent-nebius-auth-diagnose
Diagnose Nebius project authentication, quota reads, and selector failures without mutation.
nebiussdlc-validate-codes
Validates implemented features with build, lint, type, config, and review-only code-review checks.
Frequently Asked Questions About Nebius
FAQPage SchemaWhat specific tasks can be performed using Nebius infrastructure?▼
Nebius enables the deployment of Physical AI agent VMs, orchestration of multi-GPU world model training, management of robotics simulation artifacts, and execution of Sim2Real reinforcement learning pipelines using declarative configuration files.
Who is the target persona for these technical capabilities?▼
The platform is designed for robotics engineers, machine learning researchers, and infrastructure architects focused on high-performance computing, GPU-accelerated simulation, and the lifecycle management of physical-world models.
What are the primary prerequisites for running workloads on Nebius?▼
Users require access to Nebius cloud resources, configured IAM authentication, and familiarity with declarative YAML specifications for defining cluster jobs, containerized services, and multi-stage pipeline execution.