@netlify/agent-runner-cli 1.198.0-PAW-740.0 → 1.199.0

package/dist/skills/netlify-ai-gateway/SKILL.md

@@ -27,6 +27,14 @@ Gemini variables independently.
 
 AI Gateway requires a **credit-based plan** (Free, Personal, or Pro). It is not available on legacy pricing plans.
 
+### Rules
+
+- Use the provider SDK with a default constructor (`new Anthropic()`, `new OpenAI()`, `new GoogleGenAI({})`). Don't
+  pass `apiKey` or `baseURL`, except in the Image Generation example below.
+- Don't write checks that return an error when an API key variable is missing, and don't ask the user for an API key.
+  Netlify sets the keys at runtime.
+- Use `fetch` only when no official SDK covers the call.
+
 ### Environment Variables
 
 | Provider | Variables |
@@ -35,7 +43,7 @@ AI Gateway requires a **credit-based plan** (Free, Personal, or Pro). It is not
 | **OpenAI** | `OPENAI_API_KEY`, `OPENAI_BASE_URL` |
 | **Google Gemini** | `GEMINI_API_KEY`, `GOOGLE_GEMINI_BASE_URL` |
 | **Typesafe** | `TYPESAFE_API_KEY`, `TYPESAFE_BASE_URL` |
-| **Gateway (always set)** | `NETLIFY_AI_GATEWAY_KEY`, `NETLIFY_AI_GATEWAY_BASE_URL` |
+| **Gateway (always set)** | `NETLIFY_AI_GATEWAY_KEY`, `NETLIFY_AI_GATEWAY_URL` |
 
 ## Quick Start
 
@@ -228,7 +236,7 @@ import { GoogleGenAI } from '@google/genai'
 
 const ai = new GoogleGenAI({
   apiKey: process.env.NETLIFY_AI_GATEWAY_KEY,
-  httpOptions: { baseUrl: process.env.NETLIFY_AI_GATEWAY_BASE_URL?.replace(/\/$/, '') },
+  httpOptions: { baseUrl: process.env.NETLIFY_AI_GATEWAY_URL?.replace(/\/$/, '') },
 })
 
 const response = await ai.models.generateContent({
@@ -358,10 +366,12 @@ This also works in server components, API routes in Astro, Remix loaders/actions
 
 ## REST API / Direct Fetch
 
-For HTTP-level control, use the gateway variables directly:
+Only when no official SDK covers the call. Use the gateway variables; the URL ends with a slash:
 
 ```typescript
-const response = await fetch(`${process.env.NETLIFY_AI_GATEWAY_BASE_URL}/anthropic/v1/messages`, {
+const gatewayUrl = process.env.NETLIFY_AI_GATEWAY_URL.replace(/\/$/, '')
+
+const response = await fetch(`${gatewayUrl}/anthropic/v1/messages`, {
   method: 'POST',
   headers: {
     'Content-Type': 'application/json',
@@ -378,24 +388,6 @@ const response = await fetch(`${process.env.NETLIFY_AI_GATEWAY_BASE_URL}/anthrop
 const data = await response.json()
 ```
 
-Or using the provider-specific variables with `fetch`:
-
-```typescript
-const response = await fetch(`${process.env.ANTHROPIC_BASE_URL}/v1/messages`, {
-  method: 'POST',
-  headers: {
-    'Content-Type': 'application/json',
-    'x-api-key': process.env.ANTHROPIC_API_KEY,
-    'anthropic-version': '2023-06-01',
-  },
-  body: JSON.stringify({
-    model: 'claude-opus-4-8',
-    max_tokens: 1024,
-    messages: [{ role: 'user', content: 'Hello!' }],
-  }),
-})
-```
-
 ## Available Models
 
 <!-- AVAILABLE_MODELS -->
@@ -427,13 +419,14 @@ token consumption is tracked per model and converted to Netlify credits.
 
 ### "API key not found" or env var is undefined
 
-**Cause:** AI Gateway env vars are not injected.
+**Cause:** AI Gateway env vars are not injected, or the code reads a variable the runtime doesn't set.
 
 **Fix:**
 
-1. Ensure the site has at least one production deploy
-2. Deploy to Netlify
-3. Check that you are reading the env var in server-side code, not client-side browser code
+1. Check the function uses v2 or v3 syntax, not a v1 `exports.handler`
+2. Check the code reads only the variable names in the Environment Variables table above
+3. Ensure the site has at least one production deploy
+4. Check that you are reading the env var in server-side code, not client-side browser code
 
 ### Rate limit exceeded (429)
 

package/package.json

@@ -1,7 +1,7 @@
 {
   "name": "@netlify/agent-runner-cli",
   "type": "module",
-  "version": "1.198.0-PAW-740.0",
+  "version": "1.199.0",
   "description": "CLI tool for running Netlify agents",
   "main": "./dist/index.js",
   "types": "./dist/index.d.ts",