What problem does it solve? Adding AI features to a Netlify site normally requires managing separate API keys for OpenAI, Anthropic, or Google, plus knowing which models each provider supports. This Skill explains how Netlify AI Gateway proxies those SDK calls with auto-injected credentials, which models are actually available, and how to avoid common setup errors. ## Core Features & Use Cases - Keyless SDK setup: Use the standard OpenAI, Anthropic, and @google/genai SDKs with auto-injected base URLs and placeholder keys, no provider accounts needed. - Curated model reference: Lists every chat, reasoning, and image model the gateway actually supports, preventing "model not found" errors from unsupported model names. - Image generation guidance: Shows text-to-image and image-to-image patterns using Gemini image models, the only image provider routed through the gateway. - Use Case: You are building a Netlify Function that generates product descriptions. Use this Skill to wire up the OpenAI SDK with zero key configuration, pick a supported model like gpt-4o-mini, and debug why local dev fails before the first production deploy. ## Quick Start Ask the assistant to add an AI chat endpoint to your Netlify site using the AI Gateway with a supported model.
