Tenzro Cloud SDK

Official TypeScript SDK for Tenzro Cloud - AI-Native Cloud Infrastructure.

Installation

npm install @tenzro/cloud

Quick Start

import { Tenzro } from '@tenzro/cloud';

const client = new Tenzro({
  apiKey: process.env.TENZRO_API_KEY!,
});

// AI Inference
const response = await client.ai.infer({
  model: 'gemini-3-flash',
  messages: [{ role: 'user', content: 'Hello!' }],
});

console.log(response.content);

Services

| Service | Description | |---------|-------------| | client.ai | AI inference | | client.agents | AI agents | | client.workflows | Workflow automation | | client.vec | Vector database | | client.kev | Key-value store | | client.data | SQL database | | client.graph | Graph database | | client.files | Object storage | | client.hub | Model registry | | client.security | Key management | | client.enclaves | Confidential computing (TEE) | | client.server | MCP server | | client.cortex | Tiny models and training | | client.projects | Project management | | client.apiKeys | API key management |

AI Inference

Inference Endpoints

// Create reusable endpoint
const endpoint = await client.ai.createEndpoint({
  endpointName: 'customer-support',
  model: 'gemini-3-flash',
  systemPrompt: 'You are a helpful assistant.',
  temperature: 0.7,
});

// Run inference
const response = await client.ai.inferWithEndpoint(endpoint.endpointId, {
  prompt: 'How do I reset my password?',
});

Embeddings

const result = await client.ai.embed({
  texts: ['Hello world', 'Machine learning'],
  taskType: 'RETRIEVAL_DOCUMENT',
  outputDimensionality: 768,
});

Agents

const agent = await client.agents.create({
  agentName: 'Research Assistant',
  endpointId: endpoint.endpointId,
  systemPrompt: 'You help users research topics.',
  orchestrationPattern: 'SINGLE',
});

await client.agents.activate(agent.agentId);

const result = await client.agents.chat(agent.agentId, {
  message: 'Find information about renewable energy',
});

Orchestration Patterns

• SINGLE - Single agent processing
• SUPERVISOR - Supervisor delegates to sub-agents
• ROUTER - Routes to specialized agents
• HIERARCHICAL - Multi-level agent hierarchy
• SWARM - Collaborative agent swarm
• SEQUENTIAL - Chain of agents
• PARALLEL - Concurrent execution

Vector Database

const db = await client.vec.createDatabase({
  dbName: 'documents',
  dimension: 768,
  metricType: 'COSINE',
});

await client.vec.insert(db.vecDbId, {
  vectors: [{
    id: 'doc1',
    vector: embedding,
    metadata: { title: 'Document 1' },
  }],
});

const results = await client.vec.search(db.vecDbId, {
  vector: queryEmbedding,
  topK: 10,
});

SQL Database

const db = await client.data.createDatabase({ dbName: 'myapp' });

await client.data.createTableFromSchema(db.dataDbId, 'users', {
  id: { type: 'int', primaryKey: true, autoIncrement: true },
  name: { type: 'string', nullable: false },
  email: { type: 'string', unique: true },
});

const results = await client.data.from(db.dataDbId, 'users')
  .select('id', 'name', 'email')
  .where('name', '=', 'John')
  .execute();

MCP Server

const server = await client.server.create({
  deploymentName: 'knowledge-assistant',
  endpointId: endpoint.endpointId,
  dataSources: [
    { type: 'VEC', id: vecDbId, name: 'knowledge' },
    { type: 'DATA', id: dataDbId, name: 'business' },
  ],
});

const response = await client.server.chat(server.deploymentId, {
  message: 'What products do we have?',
});

Security

const key = await client.security.createKey({
  keyName: 'data-encryption',
  purpose: 'ENCRYPT_DECRYPT',
  algorithm: 'AES_256_GCM',
});

const encrypted = await client.security.encrypt({
  keyId: key.keyId,
  data: 'sensitive information',
});

const decrypted = await client.security.decrypt({
  keyId: key.keyId,
  data: encrypted.ciphertext,
});

Error Handling

import { APIError, AuthenticationError, RateLimitError } from '@tenzro/cloud';

try {
  const response = await client.ai.infer({ ... });
} catch (error) {
  if (error instanceof AuthenticationError) {
    console.error('Invalid API key');
  } else if (error instanceof RateLimitError) {
    console.error(`Rate limited. Retry after ${error.retryAfter}s`);
  } else if (error instanceof APIError) {
    console.error(`API error: ${error.message}`);
  }
}

Configuration

const client = new Tenzro({
  apiKey: 'sk_xxx',
  projectId: 'proj_xxx',          // Auto-injected into all requests
  ai: {                            // Default AI configuration
    model: 'gemini-3-flash',
    temperature: 0.7,
  },
  timeout: 60000,
  maxRetries: 2,
});

Requirements

• Node.js 22+
• TypeScript 5.0+ (recommended)

Links

• Console
• Documentation
• Website

License

Apache-2.0 - Tenzro, Inc.