Quick Start Guide

Get up and running with Veriprompt in 5 minutes. This guide will walk you through making your first API call and seeing Veriprompt in action.

Step 1: Get Your API Key

Option A: Using Our Cloud Service

Sign up at veriprompt.com
Navigate to Developer Portal in your dashboard
Click "Generate API Key"
Copy your API key (starts with vp_)

Option B: Self-Hosted

If you're running Veriprompt on your own infrastructure:

Log in to your Veriprompt instance
Navigate to Settings → API Keys
Generate a new API key
Copy and store it securely

Keep Your API Key Secret

Never commit API keys to version control or share them publicly. Store them as environment variables.

Step 2: Make Your First API Call

Choose your preferred language:

cURLJavaScriptPython

bash

curl -X POST https://app.veriprompt.tech/api/gateway/execute \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "prompt": "Explain quantum computing in simple terms",
    "provider": "auto",
    "maxTokens": 150
  }'

javascript

// Install: npm install @veriprompt/sdk
import { VeripromptClient } from '@veriprompt/sdk';

const client = new VeripromptClient({
  apiKey: process.env.VERIPROMPT_API_KEY
});

async function runPrompt() {
  const response = await client.execute({
    prompt: 'Explain quantum computing in simple terms',
    provider: 'auto', // Let Veriprompt choose the best provider
    maxTokens: 150
  });

  console.log(response.text);
  console.log(`Used provider: ${response.provider}`);
  console.log(`Cost: $${response.cost}`);
}

runPrompt();

python

# Install: pip install veriprompt
from veriprompt import VeripromptClient

client = VeripromptClient(api_key="YOUR_API_KEY")

response = client.execute(
    prompt="Explain quantum computing in simple terms",
    provider="auto",  # Let Veriprompt choose the best provider
    max_tokens=150
)

print(response.text)
print(f"Used provider: {response.provider}")
print(f"Cost: ${response.cost}")

Pro Tip

Use provider: "auto" to let Veriprompt's intelligent routing choose the optimal provider based on cost, performance, and quality!

Step 3: Understand the Response

You'll receive a response like this:

json

{
  "text": "Quantum computing harnesses quantum mechanics to process information in fundamentally new ways...",
  "provider": "anthropic",
  "model": "claude-3-sonnet",
  "tokens": {
    "prompt": 8,
    "completion": 142,
    "total": 150
  },
  "cost": 0.0012,
  "latency": 1234,
  "requestId": "req_abc123xyz",
  "timestamp": "2024-11-10T10:30:00Z"
}

Key Response Fields

Field	Description
`text`	The AI-generated response
`provider`	Which AI provider was used (e.g., `anthropic`, `openai`)
`model`	The specific model used
`tokens`	Token usage breakdown
`cost`	Actual cost in USD for this request
`latency`	Response time in milliseconds
`requestId`	Unique ID for tracking and debugging

Step 4: Add Security Scanning

Protect against prompt injection attacks with built-in security:

JavaScriptPythoncURL

javascript

const response = await client.execute({
  prompt: userInput, // User-provided input
  provider: 'auto',
  security: {
    scanInjection: true,      // Detect injection attempts
    blockOnThreat: true,       // Block if threats detected
    minThreatScore: 0.7       // Threat threshold (0-1)
  }
});

if (response.security.threatsDetected) {
  console.error('Security threat detected:', response.security.threats);
  // Handle the threat appropriately
} else {
  console.log(response.text);
}

python

response = client.execute(
    prompt=user_input,  # User-provided input
    provider="auto",
    security={
        "scan_injection": True,      # Detect injection attempts
        "block_on_threat": True,     # Block if threats detected
        "min_threat_score": 0.7      # Threat threshold (0-1)
    }
)

if response.security.threats_detected:
    print("Security threat detected:", response.security.threats)
    # Handle the threat appropriately
else:
    print(response.text)

bash

curl -X POST https://app.veriprompt.tech/api/gateway/execute \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "prompt": "User input here",
    "provider": "auto",
    "security": {
      "scanInjection": true,
      "blockOnThreat": true,
      "minThreatScore": 0.7
    }
  }'

Step 5: Optimize for Cost

Route to the most cost-effective provider:

JavaScriptPython

javascript

const response = await client.execute({
  prompt: 'Summarize this article...',
  routing: {
    strategy: 'cost-optimized',   // Prioritize lowest cost
    maxCost: 0.001,               // Max cost per request
    minQuality: 80                // Minimum quality score (0-100)
  }
});

console.log(`Saved money! Used ${response.provider} at $${response.cost}`);

python

response = client.execute(
    prompt="Summarize this article...",
    routing={
        "strategy": "cost-optimized",  # Prioritize lowest cost
        "max_cost": 0.001,            # Max cost per request
        "min_quality": 80             # Minimum quality score (0-100)
    }
)

print(f"Saved money! Used {response.provider} at ${response.cost}")

Result: Veriprompt automatically routes to providers like DeepSeek or smaller models when appropriate, saving you money without sacrificing quality.

Step 6: Use Stored Prompts

Store prompts in the repository for versioning and reuse:

Create a Prompt in the UI

Navigate to Prompt Repository
Click "New Prompt"
Enter prompt details:
- ID: welcome-message
- System Prompt: You are a friendly customer service assistant
- User Prompt: Greet the customer and ask how you can help
Click Save

Execute the Stored Prompt

JavaScriptPythoncURL

javascript

const response = await client.executePrompt({
  promptId: 'welcome-message',
  variables: {
    customerName: 'John'
  }
});

python

response = client.execute_prompt(
    prompt_id="welcome-message",
    variables={
        "customer_name": "John"
    }
)

bash

curl -X POST https://app.veriprompt.tech/api/gateway/execute \
  -H "Content-Type: application/json" \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -d '{
    "promptId": "welcome-message",
    "variables": {
      "customerName": "John"
    }
  }'

Common Use Cases

1. Customer Support Chatbot

javascript

async function handleCustomerQuery(userMessage, conversationHistory) {
  const response = await client.execute({
    prompt: userMessage,
    context: conversationHistory,
    routing: {
      strategy: 'balanced',  // Balance cost and quality
      maxLatency: 2000       // Max 2 second response
    },
    security: {
      scanInjection: true    // Protect against prompt injection
    }
  });

  return response.text;
}

2. Content Generation

javascript

async function generateBlogPost(topic, tone) {
  const response = await client.execute({
    prompt: `Write a blog post about ${topic} in a ${tone} tone`,
    routing: {
      strategy: 'quality-first',  // Prioritize quality
      minQuality: 90
    },
    maxTokens: 1000
  });

  return response.text;
}

3. Data Analysis

javascript

async function analyzeData(csvData) {
  const response = await client.execute({
    prompt: `Analyze this data and provide insights: ${csvData}`,
    provider: 'claude-3-opus',  // Use specific high-capability model
    attachments: [{
      type: 'text/csv',
      content: csvData
    }]
  });

  return response.text;
}

Next Steps

🔐 Authentication

Learn about API keys, OAuth, and security best practices

Learn more →

📚 API Reference

Complete endpoint documentation with all parameters

Explore API →

🎯 Routing Strategies

Master intelligent routing for cost and performance

Deep dive →

🔒 Security Features

Protect your application from prompt injection and more

Security guide →

Troubleshooting

"Unauthorized" Error

Make sure your API key is:

Valid and not expired
Included in the Authorization: Bearer header
Not accidentally shared or committed to GitHub

"Rate Limited" Error

You've exceeded your tier's rate limit. Solutions:

Implement exponential backoff
Upgrade your tier
Use caching for repeated requests

"Provider Unavailable" Error

The selected provider is temporarily down. Solutions:

Use provider: "auto" for automatic failover
Configure fallback providers in routing policy
Check status.veriprompt.com

View Complete Troubleshooting Guide →

Get Help

📚 Full API Documentation
❓ FAQ
💬 Community Forum
📧 support@veriprompt.com

Congratulations! 🎉 You've made your first API call with Veriprompt. Ready to build something amazing?

Explore User Guides → | View API Reference → | Read FAQ →

Quick Start Guide ​

Step 1: Get Your API Key ​

Option A: Using Our Cloud Service ​

Option B: Self-Hosted ​

Step 2: Make Your First API Call ​

Step 3: Understand the Response ​

Key Response Fields ​

Step 4: Add Security Scanning ​

Step 5: Optimize for Cost ​

Step 6: Use Stored Prompts ​

Create a Prompt in the UI ​

Execute the Stored Prompt ​

Common Use Cases ​

1. Customer Support Chatbot ​

2. Content Generation ​

3. Data Analysis ​

Next Steps ​

🔐 Authentication

📚 API Reference

🎯 Routing Strategies

🔒 Security Features

Troubleshooting ​

"Unauthorized" Error ​

"Rate Limited" Error ​

"Provider Unavailable" Error ​

Get Help ​

Quick Start Guide

Step 1: Get Your API Key

Option A: Using Our Cloud Service

Option B: Self-Hosted

Step 2: Make Your First API Call

Step 3: Understand the Response

Key Response Fields

Step 4: Add Security Scanning

Step 5: Optimize for Cost

Step 6: Use Stored Prompts

Create a Prompt in the UI

Execute the Stored Prompt

Common Use Cases

1. Customer Support Chatbot

2. Content Generation

3. Data Analysis

Next Steps

Troubleshooting

"Unauthorized" Error

"Rate Limited" Error

"Provider Unavailable" Error

Get Help