@fetch-hive/sdk

Official Node.js / TypeScript SDK for Fetch Hive — invoke AI prompts, workflows, and agents from your application.

Installation

npm install @fetch-hive/sdk
# or
yarn add @fetch-hive/sdk
# or
pnpm add @fetch-hive/sdk

Quick start

import { FetchHive } from '@fetch-hive/sdk';

const client = new FetchHive({ apiKey: process.env.FETCH_HIVE_API_KEY! });

Get your API key from the Fetch Hive dashboard.

Invoke a prompt

const result = await client.invokePrompt({
  deployment: 'my-prompt',
  inputs: { name: 'Alice', topic: 'machine learning' },
  metadata: {},
});
console.log(result.response);

Invoke a prompt (streaming)

for await (const chunk of client.invokePromptStream({
  deployment: 'my-prompt',
  inputs: { name: 'Alice' },
})) {
  if (chunk.type === 'response') process.stdout.write(chunk.response ?? '');
  if (chunk.type === 'usage')    console.log('\nUsage:', chunk.usage);
}

Invoke a workflow

const run = await client.invokeWorkflow({
  deployment: 'my-workflow',
  inputs: { customer_id: '42' },
  metadata: {},
});
console.log(run.status, run.output);

Invoke a workflow (async)

const run = await client.invokeWorkflow({
  deployment: 'my-workflow',
  inputs: { customer_id: '42' },
  async: { enabled: true, callback_url: 'https://example.com/webhook' },
});
console.log('Queued:', run.run_id);

Invoke an agent

const reply = await client.invokeAgent({
  agent: 'my-agent',
  message: 'What is the weather in London?',
  metadata: {},
});
console.log(reply.response);

Metadata

Pass optional metadata on prompt, workflow, or agent invokes to attach flat audit fields for log display and filtering. Metadata values must be strings, numbers, booleans, or null.

Invoke an agent (streaming)

for await (const chunk of client.invokeAgentStream({
  agent: 'my-agent',
  message: 'What is the weather in London?',
  thread_id: 'session-abc123',   // optional — persist conversation history
})) {
  if (chunk.type === 'response') process.stdout.write(chunk.response ?? '');
  if (chunk.type === 'tool')     console.log(`\nCalling tool: ${chunk.tool}`);
  if (chunk.type === 'usage')    console.log('\nUsage:', chunk.usage);
}

Multimodal (image) inputs

const result = await client.invokeAgent({
  agent: 'vision-agent',
  message: 'Describe this image',
  image_urls: ['https://example.com/photo.jpg'],
});
console.log(result.response);

Authentication

Pass the API key to the constructor:

const client = new FetchHive({ apiKey: 'fhk_...' });

Or set the environment variable and omit the explicit key:

export FETCH_HIVE_API_KEY=fhk_...

Configuration

| Option | Default | Description | |---|---|---| | apiKey | process.env.FETCH_HIVE_API_KEY | Bearer token from the Fetch Hive dashboard | | baseURL | https://api.fetchhive.com/v1 | Override the API base URL |

Links

• Fetch Hive dashboard
• API documentation
• GitHub

Version

0.3.0

License

MIT — see LICENSE.