develop-ai-functions-example

Name: develop-ai-functions-example
Author: vercel

1.4kinstalls4.8kforks

How to install develop-ai-functions-example?

vercel/ai/develop-ai-functions-example

$npx -y skills add vercel/ai --skill develop-ai-functions-example

Installs into the current project.

Summary

Develop examples for AI SDK functions. Use when creating, running, or modifying examples under examples/ai-functions/src to validate provider support, demonstrate features, or create test fixtures.

Files · 1

View on GitHub

SKILL.md

1	## AI Functions Examples
2
3	The `examples/ai-functions/` directory contains scripts for validating, testing, and iterating on AI SDK functions across providers.
4
5	## Example Categories
6
7	Examples are organized by AI SDK function in `examples/ai-functions/src/`:
8
9	\| Directory \| Purpose \|
10	\| ------------------ \| ---------------------------------------------------- \|
11	\| `generate-text/` \| Non-streaming text generation with `generateText()` \|
12	\| `stream-text/` \| Streaming text generation with `streamText()` \|
13	\| `generate-object/` \| Structured output generation with `generateObject()` \|
14	\| `stream-object/` \| Streaming structured output with `streamObject()` \|
15	\| `agent/` \| `ToolLoopAgent` examples for agentic workflows \|
16	\| `embed/` \| Single embedding generation with `embed()` \|
17	\| `embed-many/` \| Batch embedding generation with `embedMany()` \|
18	\| `generate-image/` \| Image generation with `generateImage()` \|
19	\| `generate-speech/` \| Text-to-speech with `generateSpeech()` \|
20	\| `transcribe/` \| Audio transcription with `transcribe()` \|
21	\| `rerank/` \| Document reranking with `rerank()` \|
22	\| `middleware/` \| Custom middleware implementations \|
23	\| `registry/` \| Provider registry setup and usage \|
24	\| `telemetry/` \| OpenTelemetry integration \|
25	\| `complex/` \| Multi-component examples (agents, routers) \|
26	\| `lib/` \| Shared utilities (not examples) \|
27	\| `tools/` \| Reusable tool definitions \|
28
29	## File Naming Convention
30
31	Examples follow the pattern: `{provider}-{feature}.ts`
32
33	\| Pattern \| Example \| Description \|
34	\| ---------------------------------------- \| ------------------------------------------ \| -------------------------- \|
35	\| `{provider}.ts` \| `openai.ts` \| Basic provider usage \|
36	\| `{provider}-{feature}.ts` \| `openai-tool-call.ts` \| Specific feature \|
37	\| `{provider}-{sub-provider}.ts` \| `amazon-bedrock-anthropic.ts` \| Provider with sub-provider \|
38	\| `{provider}-{sub-provider}-{feature}.ts` \| `google-vertex-anthropic-cache-control.ts` \| Sub-provider with feature \|
39
40	## Example Structure
41
42	All examples use the `run()` wrapper from `lib/run.ts` which:
43
44	- Loads environment variables from `.env`
45	- Provides error handling with detailed API error logging
46
47	### Basic Template
48
49	```typescript
50	import { providerName } from '@ai-sdk/provider-name';
51	import { generateText } from 'ai';
52	import { run } from '../lib/run';
53
54	run(async () => {
55	const result = await generateText({
56	model: providerName('model-id'),
57	prompt: 'Your prompt here.',
58	});
59
60	console.log(result.text);
61	console.log('Token usage:', result.usage);
62	console.log('Finish reason:', result.finishReason);
63	});
64	```
65
66	### Streaming Template
67
68	```typescript
69	import { providerName } from '@ai-sdk/provider-name';
70	import { streamText } from 'ai';
71	import { printFullStream } from '../lib/print-full-stream';
72	import { run } from '../lib/run';
73
74	run(async () => {
75	const result = streamText({
76	model: providerName('model-id'),
77	prompt: 'Your prompt here.',
78	});
79
80	await printFullStream({ result });
81	});
82	```
83
84	### Tool Calling Template
85
86	```typescript
87	import { providerName } from '@ai-sdk/provider-name';
88	import { generateText, tool } from 'ai';
89	import { z } from 'zod';
90	import { run } from '../lib/run';
91
92	run(async () => {
93	const result = await generateText({
94	model: providerName('model-id'),
95	tools: {
96	myTool: tool({
97	description: 'Tool description',
98	inputSchema: z.object({
99	param: z.string().describe('Parameter description'),
100	}),
101	execute: async ({ param }) => {
102	return { result: `Processed: ${param}` };
103	},
104	}),
105	},
106	prompt: 'Use the tool to...',
107	});
108
109	console.log(JSON.stringify(result, null, 2));
110	});
111	```
112
113	### Structured Output Template
114
115	```typescript
116	import { providerName } from '@ai-sdk/provider-name';
117	import { generateObject } from 'ai';
118	import { z } from 'zod';
119	import { run } from '../lib/run';
120
121	run(async () => {
122	const result = await generateObject({
123	model: providerName('model-id'),
124	schema: z.object({
125	name: z.string(),
126	items: z.array(z.string()),
127	}),
128	prompt: 'Generate a...',
129	});
130
131	console.log(JSON.stringify(result.object, null, 2));
132	console.log

People who liked this also viewed

6 picks

Type

develop-ai-functions-example

1.4kinstalls4.8kforks

View on GitHub

How to install develop-ai-functions-example?

vercel/ai/develop-ai-functions-example

$npx -y skills add vercel/ai --skill develop-ai-functions-example

Installs into the current project.

Summary

Develop examples for AI SDK functions. Use when creating, running, or modifying examples under examples/ai-functions/src to validate provider support, demonstrate features, or create test fixtures.

Files · 1

View on GitHub

SKILL.md

1	## AI Functions Examples
2
3	The `examples/ai-functions/` directory contains scripts for validating, testing, and iterating on AI SDK functions across providers.
4
5	## Example Categories
6
7	Examples are organized by AI SDK function in `examples/ai-functions/src/`:
8
9	\| Directory \| Purpose \|
10	\| ------------------ \| ---------------------------------------------------- \|
11	\| `generate-text/` \| Non-streaming text generation with `generateText()` \|
12	\| `stream-text/` \| Streaming text generation with `streamText()` \|
13	\| `generate-object/` \| Structured output generation with `generateObject()` \|
14	\| `stream-object/` \| Streaming structured output with `streamObject()` \|
15	\| `agent/` \| `ToolLoopAgent` examples for agentic workflows \|
16	\| `embed/` \| Single embedding generation with `embed()` \|
17	\| `embed-many/` \| Batch embedding generation with `embedMany()` \|
18	\| `generate-image/` \| Image generation with `generateImage()` \|
19	\| `generate-speech/` \| Text-to-speech with `generateSpeech()` \|
20	\| `transcribe/` \| Audio transcription with `transcribe()` \|
21	\| `rerank/` \| Document reranking with `rerank()` \|
22	\| `middleware/` \| Custom middleware implementations \|
23	\| `registry/` \| Provider registry setup and usage \|
24	\| `telemetry/` \| OpenTelemetry integration \|
25	\| `complex/` \| Multi-component examples (agents, routers) \|
26	\| `lib/` \| Shared utilities (not examples) \|
27	\| `tools/` \| Reusable tool definitions \|
28
29	## File Naming Convention
30
31	Examples follow the pattern: `{provider}-{feature}.ts`
32
33	\| Pattern \| Example \| Description \|
34	\| ---------------------------------------- \| ------------------------------------------ \| -------------------------- \|
35	\| `{provider}.ts` \| `openai.ts` \| Basic provider usage \|
36	\| `{provider}-{feature}.ts` \| `openai-tool-call.ts` \| Specific feature \|
37	\| `{provider}-{sub-provider}.ts` \| `amazon-bedrock-anthropic.ts` \| Provider with sub-provider \|
38	\| `{provider}-{sub-provider}-{feature}.ts` \| `google-vertex-anthropic-cache-control.ts` \| Sub-provider with feature \|
39
40	## Example Structure
41
42	All examples use the `run()` wrapper from `lib/run.ts` which:
43
44	- Loads environment variables from `.env`
45	- Provides error handling with detailed API error logging
46
47	### Basic Template
48
49	```typescript
50	import { providerName } from '@ai-sdk/provider-name';
51	import { generateText } from 'ai';
52	import { run } from '../lib/run';
53
54	run(async () => {
55	const result = await generateText({
56	model: providerName('model-id'),
57	prompt: 'Your prompt here.',
58	});
59
60	console.log(result.text);
61	console.log('Token usage:', result.usage);
62	console.log('Finish reason:', result.finishReason);
63	});
64	```
65
66	### Streaming Template
67
68	```typescript
69	import { providerName } from '@ai-sdk/provider-name';
70	import { streamText } from 'ai';
71	import { printFullStream } from '../lib/print-full-stream';
72	import { run } from '../lib/run';
73
74	run(async () => {
75	const result = streamText({
76	model: providerName('model-id'),
77	prompt: 'Your prompt here.',
78	});
79
80	await printFullStream({ result });
81	});
82	```
83
84	### Tool Calling Template
85
86	```typescript
87	import { providerName } from '@ai-sdk/provider-name';
88	import { generateText, tool } from 'ai';
89	import { z } from 'zod';
90	import { run } from '../lib/run';
91
92	run(async () => {
93	const result = await generateText({
94	model: providerName('model-id'),
95	tools: {
96	myTool: tool({
97	description: 'Tool description',
98	inputSchema: z.object({
99	param: z.string().describe('Parameter description'),
100	}),
101	execute: async ({ param }) => {
102	return { result: `Processed: ${param}` };
103	},
104	}),
105	},
106	prompt: 'Use the tool to...',
107	});
108
109	console.log(JSON.stringify(result, null, 2));
110	});
111	```
112
113	### Structured Output Template
114
115	```typescript
116	import { providerName } from '@ai-sdk/provider-name';
117	import { generateObject } from 'ai';
118	import { z } from 'zod';
119	import { run } from '../lib/run';
120
121	run(async () => {
122	const result = await generateObject({
123	model: providerName('model-id'),
124	schema: z.object({
125	name: z.string(),
126	items: z.array(z.string()),
127	}),
128	prompt: 'Generate a...',
129	});
130
131	console.log(JSON.stringify(result.object, null, 2));
132	console.log

People who liked this also viewed

6 picks

Type