Get Expert Recommendation

Get opinionated, AI-powered starter template recommendations based on natural language queries and expert rules. This endpoint enhances the standard recommendation with expert knowledge about best practices and common misconceptions.

GET /api/recommendation/expert?question=string

Query Parameters

question

A natural language description of your requirements.

Type: string

Required: true

Example queries:

"I want a minimal React starter that I can build up from scratch"
"Looking for a backend with websockets for realtime updates"
"Need a starter with NoSQL database for flexibility"

Expert Rules

The expert recommendation system incorporates knowledge about best practices and common misconceptions in starter template selection. It can provide additional recommendations and insights that are not covered by the standard recommendation algorithm.

Response

Response Fields

question

The original query string.

extracted_query

The structured query parameters extracted by GPT.

starters

Array of matching starter templates.

recommendation

Expert recommendation explaining matches and addressing potential misconceptions.

Example Response

{
  "question": "I want a minimal React starter that I can build up from scratch",
  "extracted_query": {
    "technologies": ["React"],
    "features": ["minimal"],
    "min_quality": 3,
    "minMatched": 1,
    "limit": 3
  },
  "starters": [...],
  "recommendation": "While you might be tempted to start with a minimal template and build it up yourself, this approach often leads to reinventing the wheel and missing important best practices. Instead, I recommend the 'comprehensive-react-starter' which comes with a well-thought-out structure and essential features like testing and Docker setup. You can always remove features you don't need, but having them properly configured from the start will save you time and help you follow best practices..."
}

Error Responses

400 Bad Request

Returned when the question parameter is missing.

500 Internal Server Error

Returned when there's an error processing the request or communicating with OpenAI.