API Docs — Laravel + team collerboration

A lightweight API documentation tool built with Laravel 13 and powered by Multi model AI-assisted endpoint documentation from any json API collections.

The goal of the project was to keep things simple:

• upload a Postman collection
• browse and edit endpoints
• generate readable API documentation using local AI models

---

Features

• Import Postman Collection v2.1 files
• Create, edit, and delete endpoints from the UI
• Generate AI-written endpoint summaries using Ollama
• Stream AI responses live with SSE
• Search endpoints by name, URL, method, or summary
• Store collections locally as JSON files
• Cache generated summaries

---

Getting Started

Requirements

Dependency	Version
PHP	8.3+
Composer	Latest
Ollama	Latest

---

Install Ollama

Download Ollama:

https://ollama.com

Start the Ollama server:

ollama serve

Pull a model:

# Recommended
ollama pull llama3

# Alternatives
ollama pull mistral
ollama pull phi3
ollama pull gemma2

Suggested Models

Model	Notes
llama3	Best overall documentation quality
mistral	Faster responses
phi3	Lightweight and fast
gemma2	Good alternative option

---

Installation

Clone the repository and install dependencies:

git clone <your-repo>
cd api-docs-laravel

composer install

Copy the environment file:

cp .env.example .env
php artisan key:generate

---

Environment Configuration

Update your .env file:

OLLAMA_HOST=http://localhost:11434
OLLAMA_MODEL=llama3
OLLAMA_TEMPERATURE=0.3
OLLAMA_MAX_TOKENS=1024

APP_URL=http://localhost:8000

---

Storage Setup

Create the collections directory and link storage:

mkdir -p storage/app/collections

php artisan storage:link

---

Run the Application

php artisan serve

Application URL:

http://localhost:8000

---

API Endpoints

Collections

Method	Endpoint	Description
GET	/api/collections	Fetch all collections
POST	/api/collections/upload	Upload a Postman collection
GET	/api/collections/{id}	Fetch a collection with endpoints
PUT	/api/collections/{id}	Update collection details
DELETE	/api/collections/{id}	Remove a collection

---

Endpoints

Method	Endpoint	Description
GET	/api/collections/{id}/endpoints	List endpoints
POST	/api/collections/{id}/endpoints	Create endpoint
GET	/api/collections/{id}/endpoints/{epId}	Get endpoint
PUT	/api/collections/{id}/endpoints/{epId}	Update endpoint
DELETE	/api/collections/{id}/endpoints/{epId}	Delete endpoint

Supported filters:

?group=
?method=
?search=

---

AI Routes

Method	Endpoint	Description
GET	/api/ai/status	Check Ollama connection
GET	/api/ai/models	List installed models
POST	/api/ai/summarize	Generate docs for one endpoint
POST	/api/ai/summarize-collection	Generate docs for entire collection

---

Example AI Request

POST /api/ai/summarize

{
  "endpoint": {
    "name": "Get User",
    "method": "GET",
    "url": "{{base_url}}/api/users/:id",
    "group": "Users",
    "auth": {
      "type": "bearer"
    },
    "query": [],
    "body": {},
    "responses": []
  },
  "model": "llama3",
  "collection_id": "uuid-here",
  "endpoint_id": "uuid-here"
}

---

Project Structure

api-docs-laravel/
├── app/
│   ├── Http/
│   │   ├── Controllers/
│   │   │   ├── CollectionController.php
│   │   │   ├── EndpointController.php
│   │   │   └── OllamaController.php
│   │   └── Middleware/
│   │       └── CorsMiddleware.php
│   │
│   └── Services/
│       ├── OllamaService.php
│       └── PostmanParserService.php
│
├── bootstrap/
│   └── app.php
│
├── config/
│   └── ollama.php
│
├── resources/
│   └── views/
│       └── app.blade.php
│
├── routes/
│   ├── api.php
│   └── web.php
│
├── storage/
│   └── app/
│       └── collections/
│
└── .env.example

---

How the AI Flow Works

1. User clicks Generate AI Documentation
2. Frontend sends endpoint data to the AI route
3. Laravel builds a structured prompt
4. Ollama receives the request with streaming enabled
5. The response streams token-by-token through SSE
6. The frontend renders the response live
7. Generated summaries are saved back to the collection JSON

---

Prompt Design

The prompt includes:

• HTTP method
• Endpoint URL
• Group/folder name
• Authentication type
• Query parameters
• Path variables
• Request body
• Example responses
• Existing endpoint descriptions

Generated output follows a structure similar to Postman AI documentation:

• Overview
• Use Case
• Authentication
• Parameters
• Request Body
• Response
• Notes

---

Notes

• llama3 generally produces the cleanest documentation
• Lower temperature values work better for structured API docs
• Streaming works through Server-Sent Events (SSE)
• Collections are stored locally as JSON for simplicity

---

Future Improvements

A few things planned for later versions:

• Markdown export
• OpenAPI/Swagger generation

---

License

MIT License.