​

Chat API

​

Overview

• Vadalog code generation tailored to your specific question
• Syntax assistance and debugging help
• Documentation search integration via Algolia
• Context-aware responses based on Prometheux documentation
• Grammar-compliant code following Vadalog best practices
• Intelligent examples — simple questions get focused snippets, complex questions get full programs

​

Endpoints

​

/api/docsChat (Streaming)

• URL: https://chat-docs.prometheux.ai/api/docsChat
• Method: POST
• Content-Type: application/json
• Response: Streaming text (AI SDK format)

{
  "messages": [
    {
      "role": "user",
      "content": "Show me a PostgreSQL connection example"
    }
  ]
}

0:"Below"
0:" is"
0:" a"
0:" complete"
0:" Vadalog"
0:" example..."

const response = await fetch('https://chat-docs.prometheux.ai/api/docsChat', {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    messages: [
      { role: 'user', content: 'How do I connect to Neo4j?' }
    ]
  })
});

// Handle streaming response
const reader = response.body.getReader();
const decoder = new TextDecoder();
let result = '';

while (true) {
  const { done, value } = await reader.read();
  if (done) break;
  
  const chunk = decoder.decode(value, { stream: true });
  // Parse AI SDK format: 0:"text"
  const lines = chunk.split('\n');
  for (const line of lines) {
    if (line.startsWith('0:"') && line.endsWith('"')) {
      result += line.slice(3, -1); // Extract text content
    }
  }
}

console.log(result); // Complete Vadalog code example

import requests
import re

url = "https://chat-docs.prometheux.ai/api/docsChat"
payload = {
    "messages": [
        {"role": "user", "content": "Show me aggregation examples"}
    ]
}

response = requests.post(url, json=payload, stream=True)

result = ""
for chunk in response.iter_content(chunk_size=1024, decode_unicode=True):
    if chunk:
        # Parse AI SDK format: 0:"text"
        lines = chunk.split('\n')
        for line in lines:
            if line.startswith('0:"') and line.endswith('"'):
                result += line[3:-1]  # Extract text content

print(result)  # Complete Vadalog code

curl -X POST https://chat-docs.prometheux.ai/api/docsChat \
  -H "Content-Type: application/json" \
  -d '{
    "messages": [
      {"role": "user", "content": "How do I use math functions?"}
    ]
  }'

​

/api/vadalog (Standard JSON)

• URL: https://chat-docs.prometheux.ai/api/vadalog
• Method: POST
• Content-Type: application/json
• Response: Standard JSON

{
  "query": "Show me a PostgreSQL connection example",
  "context": "database connection",
  "include_docs": true
}

{
  "response": "Below is a complete Vadalog example...",
  "code_examples": [
    {
      "language": "vadalog",
      "code": "@bind(\"customer\", \"postgresql host=localhost port=5432 user=myuser password=mypass\", \"mydb\", \"customer\").\n@model(\"customer\", \"['id:int', 'name:string', 'age:int']\").\n@output(\"result\").",
      "description": "PostgreSQL connection with customer data processing"
    }
  ],
  "relevant_docs": [
    {
      "title": "Connecting to Databases",
      "url": "https://api.prometheux.ai/docs/learn/vadalog/data-sources",
      "excerpt": "PostgreSQL database connections..."
    }
  ],
  "metadata": {
    "provider": "Azure OpenAI",
    "model": "gpt-4o",
    "tokens_used": 1247,
    "search_results": 3
  }
}

const response = await fetch('https://chat-docs.prometheux.ai/api/vadalog', {
  method: 'POST',
  headers: {
    'Content-Type': 'application/json',
  },
  body: JSON.stringify({
    query: 'How do I use aggregations in Vadalog?',
    include_docs: true
  })
});

const data = await response.json();
console.log(data.response);        // Text response
console.log(data.code_examples);   // Extracted code blocks
console.log(data.relevant_docs);   // Related documentation

import requests

url = "https://chat-docs.prometheux.ai/api/vadalog"
payload = {
    "query": "Show me CSV file processing",
    "context": "data integration",
    "include_docs": True
}

response = requests.post(url, json=payload)
data = response.json()

print(f"Response: {data['response']}")
print(f"Code examples: {len(data['code_examples'])}")
print(f"Related docs: {len(data['relevant_docs'])}")

curl -X POST https://chat-docs.prometheux.ai/api/vadalog \
  -H "Content-Type: application/json" \
  -d '{
    "query": "Show me CSV file processing",
    "context": "data integration", 
    "include_docs": true
  }'

​

Features

• Focused, relevant code examples matched to your question
• Proper syntax with appropriate annotations (only when needed)
• Database connection examples for data source questions
• Full data processing workflows for integration scenarios
• Simple syntax examples for concept/logic questions
• Grammar-compliant code

• Uses Retrieval Augmented Generation (RAG) via Algolia
• Dynamically searches documentation based on your query
• Retrieves top 3 most relevant sections
• Always up-to-date — automatically includes new documentation
• No manual prompt updates needed when docs change

• Powered by Azure OpenAI (gpt-4o)
• Context-aware responses
• Iterative conversation support
• Syntax error prevention

​

Response Types

​

Rate Limits

​

Error Handling

{
  "error": "AI assistant not available",
  "message": "Service temporarily unavailable",
  "status": 503,
  "timestamp": "2024-01-15T10:30:00Z"
}

​

Best Practices

1. Use conversation context: Include previous messages for better responses
2. Be specific: Ask detailed questions about Vadalog syntax
3. Match question complexity:
   • For syntax questions: “How do I use the and() function?”
   • For complete programs: “Show me a complete PostgreSQL connection example”
4. Handle streaming: Implement proper streaming for real-time responses
5. Error handling: Implement retry logic for rate limits