API Reference

PRD Studio exposes a small OpenAI-compatible HTTP API. Point any client, script or agent at the base URL and send plain JSON.

Base URL

Every request starts here. Paths are appended to this host.

https://prd.haaviq.dev

No authentication header is required during the beta. Send Content-Type: application/json with a JSON body.

Quickstart

Generate a complete PRD in one request.

# generate a PRD
curl -X POST https://prd.haaviq.dev/api/generate \\
  -H "Content-Type: application/json" \\
  -d '{
    "name": "Scrapling",
    "description": "AI web scraping platform that turns any site into structured data",
    "template": "Web App"
  }'

The response contains a single markdown field with the full document.

{ "markdown": "## 1. Overview & Goals\n\n..." }

POST/api/suggest

Propose a description, feature list, target users and tech stack from an app name. Useful to pre-fill the studio form.

curl -X POST https://prd.haaviq.dev/api/suggest \\
  -H "Content-Type: application/json" \\
  -d '{ "title": "Scrapling", "template": "Web App" }'

# response
{
  "description": ["..."],
  "features":    ["...", "..."],
  "users":       ["..."],
  "tech_stack":  ["..."]
}

POST/api/generate

Return a full product requirements document covering nine sections: overview, architecture, API, database and ERD, AI prompt design, security, testing, deployment and roadmap.

curl -X POST https://prd.haaviq.dev/api/generate \\
  -H "Content-Type: application/json" \\
  -d '{
    "name": "Scrapling",
    "description": "AI web scraping platform",
    "features": "bulk scrape\nexport CSV",
    "template": "Web App"
  }'

# response
{ "markdown": "## 1. Overview & Goals ..." }

POST/api/diagram

Generate an architecture diagram (Mermaid) or an entity-relationship diagram (PlantUML). Returns both the diagram source and a rendered image URL.

curl -X POST https://prd.haaviq.dev/api/diagram \\
  -H "Content-Type: application/json" \\
  -d '{ "kind": "arch", "name": "Scrapling", "description": "..." }'

# response
{
  "kind": "arch",
  "code": "flowchart TD ...",
  "svg_url": "https://kroki.io/mermaid/svg/..."
}

GET/api/health

Check that the service is up and see which models are configured.

curl https://prd.haaviq.dev/api/health

# response
{ "ok": true, "ai_configured": true, "models": ["..."] }

Errors

Errors return a JSON object with a detail field.

{ "detail": "name and description are required" }

Rate limits

The beta is limited to 20 requests per hour per IP across all endpoints. Higher limits are available on request.