skills.rest
PKU-YuanGroup
PKU-YuanGroupOFFICIAL
Explore all skills by this author →
WebsiteGitHub

using-model-endpoint

Infer from a registered model endpoint via its native HTTP API.

288|34|Updated Jul 6, 2026
Search Tags:#HTTP API#model inference#model endpoint#API key management#scoped inference kernel
One-click install
npx skills add https://github.com/PKU-YuanGroup/OpenAI4S --skill using-model-endpoint-pku-yuangroup
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: using-model-endpoint
Source: https://github.com/PKU-YuanGroup/OpenAI4S/tree/main/skills/using-model-endpoint
Command: npx skills add https://github.com/PKU-YuanGroup/OpenAI4S --skill using-model-endpoint-pku-yuangroup

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires httpx, and includes scripts (resource) components.

What problem does it solve?

This Skill simplifies the process of making inferences using registered model endpoints, eliminating the need for complex orchestration.

Core Features & Use Cases

  • Simplified API Calls: Directly call a model's native HTTP API from a scoped inference kernel.
  • Preloaded BASE_URL: The BASE_URL for each model endpoint is preloaded, making it easy to build request URLs.
  • Scoped Network Egress: Each endpoint has its own inference kernel with network egress scoped to that endpoint.
  • Managed Endpoints: Supports managed endpoints that automatically handle model lifecycle.
  • Use Case: Imagine you have a registered model for image classification. Use this Skill to make predictions without manual setup or management.

Quick Start

Call the 'model_endpoint_name' inference endpoint to get a prediction from the input image 'input_image.png'.

Frequently Asked Questions about using-model-endpoint

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

FAQPage Schema
How do I make model inferences using an HTTP API without manual setup?▼

You can call a registered model's native HTTP API directly from a scoped inference kernel to make model inferences. This requires the model's API key and BASE_URL for request building, avoiding complex orchestration.

Can I use local and hosted endpoints for model inference?▼

Yes, scoped inference kernels support both local and hosted endpoints for model inference. You can build HTTP API requests using the endpoint's preloaded BASE_URL and API key to get predictions from either type.

What do I need to build HTTP API requests for model inference?▼

To build HTTP API requests for model inference, you need the model's API key and BASE_URL. The BASE_URL is preloaded for each registered endpoint, simplifying request URL construction within the scoped inference kernel.

How does scoped network egress work for model endpoint API calls?▼

Scoped network egress isolates network traffic by giving each model endpoint its own dedicated inference kernel. This ensures that HTTP API requests for model inference are securely routed only to their specific registered endpoint.

Does model inference with httpx require complex orchestration tools?▼

No, model inference using httpx does not require complex orchestration tools. You can directly infer using a registered model endpoint over its native HTTP API from a scoped inference kernel to get predictions with a single API call.

Related Skills

Explore More Software Engineering Skills→
glitternetworkglitternetwork
3.7k

pinme-api

Generate Worker TypeScript code for authenticated PinMe email and LLM API calls.

Official
Advanced
cabeznocabezno

outlines

Generate schema-constrained JSON, XML, and code with grammar-based token filtering.

Community
Advanced
ghufronbagaskaraghufronbagaskara

serving-llms-vllm

Deploy vLLM inference servers with OpenAI-compatible API endpoints.

Community
Advanced
dominodatalabdominodatalab
6

domino-model-endpoints

Deploy, monitor, and scale ML models as API endpoints in Domino Data Lab.

Official
Advanced