API Overview

The Autoflowly API provides programmatic access to MVP generation, diagnostics, and deployment.

Base URL

https://api.autoflowly.com/api

All API requests must be made over HTTPS.

Authentication

Include your JWT token in the request header:

curl -H "Authorization: Bearer YOUR_JWT_TOKEN" \
  https://api.autoflowly.com/api/conversational-mvp/quick

This is the same backend the web app and the MCP integration use — if you want ChatGPT, Claude, or Gemini to build/operate apps on your behalf without writing any HTTP calls yourself, use the MCP server instead of this API directly.

Core Endpoints

App Generation

Method	Endpoint	Description
`POST`	`/conversational-mvp/quick`	Generate a full-stack app from a description, with smart defaults
`POST`	`/conversational-mvp/chat`	Conversational generation with follow-ups
`GET`	`/conversational-mvp/session/{id}`	Get session state and generated files

Editor & SmartDoctor

Method	Endpoint	Description
`POST`	`/conversational-mvp/editor/doctor/diagnose`	Diagnose issues in an app
`POST`	`/conversational-mvp/editor/doctor/fix`	Auto-fix all detected issues
`POST`	`/conversational-mvp/editor/save-file`	Save a file change

App Branching

Method	Endpoint	Description
`POST`	`/conversational-mvp/{mvp_id}/branches`	Fork the app's current files into a named branch
`GET`	`/conversational-mvp/{mvp_id}/branches`	List an app's branches
`POST`	`/conversational-mvp/{mvp_id}/branches/{branch_id}/activate`	Open a branch's live preview
`DELETE`	`/conversational-mvp/{mvp_id}/branches/{branch_id}`	Discard a branch

Quick Start Example

Generate an app with a single API call:

const response = await fetch('https://api.autoflowly.com/api/conversational-mvp/quick', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_JWT_TOKEN',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    app_name: "Budget Tracker",
    description: "Personal budget tracker for freelancers with expense categories, income tracking, and monthly reports"
  })
})

const data = await response.json()
console.log(data.preview_url)  // https://mvp-budget-tracker.preview.autoflowly.com

Diagnose and Fix

Run SmartDoctor diagnostics on a generated MVP:

// Step 1: Diagnose
const diagnosis = await fetch('https://api.autoflowly.com/api/conversational-mvp/editor/doctor/diagnose', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_JWT_TOKEN',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    session_id: "your-session-id",
    mvp_id: "your-mvp-id"
  })
})

const report = await diagnosis.json()
// report.health_score: 85
// report.issues: [{ severity: "warning", title: "Missing dependency", ... }]

// Step 2: Fix all issues
const fix = await fetch('https://api.autoflowly.com/api/conversational-mvp/editor/doctor/fix', {
  method: 'POST',
  headers: {
    'Authorization': 'Bearer YOUR_JWT_TOKEN',
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    session_id: "your-session-id",
    mvp_id: "your-mvp-id",
    auto_redeploy: true,
    create_snapshot: true
  })
})

AI Providers

Generation uses Claude by default and automatically falls back to OpenAI if the primary provider is unavailable or errors — this is handled for you, no provider selection required.

Rate Limits

Generation and editor endpoints are rate-limited per account to prevent abuse; limits scale with your plan. If you hit one, the response is a 429 with a Retry-After header.

Error Responses

All errors follow a consistent format:

{
  "detail": "Description of the error",
  "status_code": 400
}

Code	Meaning
`400`	Invalid request parameters
`401`	Missing or invalid authentication
`404`	Resource not found
`429`	Rate limit exceeded
`500`	Internal server error