Use this skill as the top-level entrypoint for user-facing NemoClaw workflows.
The detailed workflow prompts are bundled under references/ and indexed in
skills.md.
Routing
Read the relevant referenced skill prompt before answering:
- Quickstart, prerequisites, Windows setup, first sandbox:
references/nemoclaw-user-get-started/SKILL.md - Overview, ecosystem, release notes, OpenClaw/OpenShell/NemoClaw relationship:
references/nemoclaw-user-overview/SKILL.md - Architecture, CLI selection, command reference, troubleshooting:
references/nemoclaw-user-reference/SKILL.md - Local inference, Ollama, vLLM, TensorRT-LLM, NIM, OpenAI-compatible endpoints:
references/nemoclaw-user-configure-inference/SKILL.md - Security posture, sandbox controls, credential storage, OpenClaw controls:
references/nemoclaw-user-configure-security/SKILL.md - Remote GPU or Brev deployment:
references/nemoclaw-user-deploy-remote/SKILL.md - Sandbox lifecycle, status, logs, rebuilds, upgrades, uninstall, messaging, persistence:
references/nemoclaw-user-manage-sandboxes/SKILL.md - Network policy, egress rules, endpoint access, approval workflow:
references/nemoclaw-user-manage-policy/SKILL.md - Monitoring, logs, health checks, debugging agent behavior:
references/nemoclaw-user-monitor-sandbox/SKILL.md - NemoClaw agent skills and coding assistant integration:
references/nemoclaw-user-agent-skills/SKILL.md - AI-assisted issue and PR triage instructions:
references/nemoclaw-user-triage-instructions/SKILL.md
If multiple areas apply, start with the highest-level overview or reference skill, then load more specific prompts as needed. Preserve the source skill's commands, cautions, and prerequisites when composing the final guidance.