Unite for developers and agents
Public API, MCP server, NLWeb and agent files for the Unite website. Read-only, no authentication.
The public content of the Unite website (solutions, FAQ and contact) is also available as data for AI agents and integrations. Everything is public, read-only and unauthenticated: no sign-up, API key or OAuth. The Unite platform (https://myunite.ai/) has its own login and is not part of this API.
Quickstart
curl --fail 'https://unitebr.com/api/v1/busca?query=pre-test&limit=3&lang=en'API REST (v1)
- OpenAPI 3.1 specification: https://unitebr.com/openapi.json
GET /api/v1/solucoes: solutions (query,limit,cursorfilters)GET /api/v1/solucoes/{id}: one solution (create,optimize,agente,validate,brandlocal)GET /api/v1/faq: frequently asked questions (query,limit,cursorfilters)GET /api/v1/busca?query=: natural-language search, schema.org itemsGET /api/v1/contato: contact channels and company detailsGET /api/v1/health: status
Language: lang=pt (default), lang=en or lang=es on solutions, FAQ, search and contact. Responses include Content-Language.
Cursor pagination: follow next_cursor until it is null. Responses carry an ETag (use If-None-Match). Errors are application/problem+json with status, title, detail and instance: 400 for an invalid parameter, 404 for a missing resource and 405 for an unsupported method. CORS is open for reads.
MCP server
- Endpoint (Streamable HTTP): https://unitebr.com/mcp
- Server card: https://unitebr.com/.well-known/mcp/server-card.json
- Tools:
buscar_conteudo,listar_solucoes,obter_solucao,listar_faq,obter_contato,ler_pagina. All read-only, with an optionalidiomaargument (pt, en or es). - MCP App: the search and listing tools open a view (
ui://unite/explorer.html) in compatible clients.
Sample MCP client configuration:
{
"mcpServers": {
"unite": {
"type": "http",
"url": "https://unitebr.com/mcp"
}
}
}NLWeb and structured data
POST https://unitebr.com/askwith{"query":{"text":"..."}}(NLWeb v0.55);"prefer":{"streaming":true}streams over SSE.- Schema Feeds: https://unitebr.com/schemamap.xml → https://unitebr.com/feeds/schema.jsonl.
Files for agents
- llms.txt and llms-full.txt
- In English and Spanish: en/llms.txt, es/llms.txt
- agents.md
- ARD, AI Catalog and API Catalog (RFC 9727)
- Authentication and API policy
Support
Questions about the API: support@myunite.ai. To see the platform, book a demo at https://unitebr.com/en/contact.