TLDR - copy/paste;

for AI Agents (like n8n, make…)

Method: POST

Base URL: https://api.tropicalia.dev/api/v1/agents/query

Headers

Authorization

Bearer $your_Tropicalia_API_Key

Content-Type

application/json

API Key: get in https://app.tropicalia.dev/api-keys

Project ID (project-id-string): get in https://app.tropicalia.dev/projects

Body

{
	"query": "USER-QUERY",
	"datastore_id": "project-id",
	"session_id": "OPTIONAL-session-id"
}

Vibe Coding Prompt Example (Lovable, Windsurf, Base44…):

Create a Q&A chatbot and use as information source the following configuration:
Consult your documents using AI-powered RAG (Retrieval-Augmented Generation) capabilities.

Create a backend proxy that:
- Accepts requests from your frontend
- Makes the actual API call to Tropicalia (server-to-server, no CORS issues)
- Returns the response to your frontend

Method: POST
BaseURL: <https://api.tropicalia.dev/api/v1/agents/query> 
API Key: tr_dDHOHM-eMZ-RPepdzIRICOGpJGmW5Cl651kE4IDR
Datastore_id: e578e853-a987-4ccd-ae28-226169a01ad9

Body:
{
	"query": "USER-QUERY",
	"datastore_id": "project-id",
	"session_id": "OPTIONAL-session-id"
}

Response:
{
	"query": "Quais são as principais descobertas sobre satisfação do cliente?",
	"response": "Com base nos documentos, a satisfação do cliente aumentou 15% devido à melhoria nos tempos de resposta...",
	"citations": 
	[
	{
		"index": 1,
		"document_id": "project-id-string",
		"content": "Resultados da pesquisa de satisfação do cliente mostram...",
		"original_filename": "search-result.pdf",
		"start_char": 1250,
		"end_char": 1580
	}
	],
	"processing_status": "completed"
}