skills.rest
sayalinvidia
sayalinvidiaCOMMUNITY
Explore all skills by this author →
GitHub

nemoclaw-user-configure-inference

Configure NemoClaw to route sandboxed inference to local OpenAI-compatible endpoints.

Updated Apr 16, 2026
Search Tags:#sub-agents#ollama#vllm#local-server#nemoclaw#inference-routing
One-click install
npx skills add https://github.com/sayalinvidia/sayali-skills-test --skill nemoclaw-user-configure-inference-sayalinvidia
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: nemoclaw-user-configure-inference
Source: https://github.com/sayalinvidia/sayali-skills-test/tree/main/skills/nemoclaw-user-configure-inference
Command: npx skills add https://github.com/sayalinvidia/sayali-skills-test --skill nemoclaw-user-configure-inference-sayalinvidia

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

NemoClaw provides the ability to route agent inference to local servers (Ollama, vLLM, TensorRT-LLM, NIM) or compatible endpoints while keeping secrets on the host, enabling private and low-latency workflows.

Core Features & Use Cases

  • Route to local inference servers (Ollama, vLLM, NIM) and OpenAI-compatible endpoints without exposing API keys.
  • Validate active provider and model, switch providers at runtime, and update sandbox config without full re-onboarding.
  • Manage per-agent workspace and credentials for auxiliary sub-agents and nested routing scenarios.

Quick Start

Configure NemoClaw to route the sandbox's inference.local route at your chosen local provider and verify the active route.

Frequently Asked Questions about nemoclaw-user-configure-inference

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I route local model inference to Ollama or vLLM without exposing API keys?▼

Route local model inference by configuring NemoClaw to direct sandboxed requests to local servers like Ollama or vLLM. This setup keeps credentials on the host, enabling private, low-latency workflows without exposing secrets.

Can I switch inference providers at runtime without completing full re-onboarding?▼

Yes, you can switch inference providers at runtime. NemoClaw allows you to validate active providers, update sandbox configurations, and change models dynamically without requiring a full re-onboarding process.

Does NemoClaw support routing inference for sub-agents to local servers?▼

Yes, NemoClaw supports sub-agent inference routing to local servers. It manages per-agent workspaces and credentials for auxiliary sub-agents, handling nested routing scenarios through OpenClaw.

How do I validate OpenShell gateway routing for local inference servers?▼

Validate OpenShell gateway routing by applying NemoClaw's configuration checks. This ensures correct routing to local servers like TensorRT-LLM or NIM and verifies provider validation without exposing host secrets.

What local inference servers are compatible with NemoClaw routing?▼

NemoClaw routing is compatible with Ollama, vLLM, TensorRT-LLM, NIM, and other OpenAI-compatible endpoints. This allows flexible, private, and low-latency model inference workflows.

Why does my local inference routing fail when switching providers in a sandbox?▼

Local inference routing may fail if the OpenShell gateway validation is incorrect. Use NemoClaw's troubleshooting documentation to verify provider validation, update sandbox configs, and ensure credentials remain on the host.

Related Skills

Explore More Software Engineering Skills→
gabergergaberger
4

hex-inference

Configure inference providers and map local models to hex agent tiers.

Community
Advanced
yo-stevenyo-steven

nemoclaw-user-get-started

Install NemoClaw, create a sandbox, and configure an OpenClaw inference provider.

Community
Intermediate
jetsonhacksjetsonhacks
26

nemoclaw-debug-inference-path

Diagnose NemoClaw gateway versus embedded inference routing on Jetson Orin.

Community
Advanced
overviewlabsoverviewlabs

serving-llms-vllm

Deploy and scale LLM inference with vLLM for low-latency production serving.

Community
Advanced