What is llms.txt? How NewsPro Supports AI Assistants and LLM Integrations

What Is llms.txt — and Why Does It Matter?

In 2024, a new standard quietly emerged that is changing how AI assistants understand software platforms: llms.txt. Just like robots.txt tells search engine crawlers how to index your site, llms.txt tells large language models (LLMs) like ChatGPT, Claude, Gemini, and Perplexity how to understand your platform, your API, and your capabilities.

When someone asks an AI assistant "How do I create an article using the NewsPro API?" — the AI can now give an accurate, up-to-date answer. Because NewsPro provides a structured documentation file designed specifically for AI consumption.

NewsPro's llms.txt and llms-full.txt

NewsPro publishes two AI-readable documentation files:

• newspro.io/llms.txt — A concise summary of the platform: what NewsPro is, what APIs are available, and where to find more information. Ideal for quick context injection.
• newspro.io/llms-full.txt — The complete API reference in a single flat file. Every resource, every action, every parameter — all in one place, formatted for LLM consumption.

These files follow the llms.txt open standard, which is gaining rapid adoption across developer tools, SaaS platforms, and documentation sites.

The NewsPro REST API: Every Resource Explained

All NewsPro API requests share the same base endpoint and authentication method:

• Endpoint: POST https://yoursite.newspro.io/functions/api
• Authentication: Authorization: Bearer YOUR_SECRET_KEY
• Format: JSON body with at minimum resource and action fields

Every resource follows consistent CRUD patterns using four standard actions: get, add, edit, delete. Some resources also expose specialized actions for advanced use cases.

Articles API

The core of any news site. The Articles API gives you full programmatic control over your content — create, retrieve, update, and delete articles with over 25 available fields:

• Content: title, slug, subtitle, body (HTML)
• Organization: category_id, country_id, type_id, user_id (author)
• Status: status (1=published, 0=draft), approved (1=approved, 0=pending)
• Media: thumbnail (media file ID), video_link
• Flags: sponsored, anonymous, display_article_home, votes
• Analytics: total_views, likes, dislikes, comments
• Filters: category_id, country_id, user_id, approved, status, date_from, date_to, search

Categories API

Manage the full category hierarchy of your site — including nested categories, navigation placement, and visual styling:

• Core: title, slug, description, color (hex)
• Hierarchy: parent_id (0 = top-level)
• Navigation: category_top, top_menu_position, hamburger_menu_position
• Status: status (3=active, 0=inactive), approved
• Filters: parent_id, status, approved, search

Pages API

Create and manage static pages — About, Contact, Terms, Privacy, or any custom page — separate from the article feed:

• Core: title, slug, description (meta), content (full HTML)
• Media: thumbnail (media file ID)
• Status: status (1=published, 0=draft)
• Analytics: total_views
• Filters: status, search

Breaking News API

Programmatically push urgent alerts to your site's breaking news ticker — great for automated monitoring feeds or external integrations:

• Core: title (headline), content (additional text or URL)
• Attribution: user_id
• Status: status (1=active, 0=inactive)
• Filters: user_id, status, date_from, date_to

Users API

Full account management across all four user roles — admin, editor, author, and subscriber. Note: uses user_id instead of id for lookups:

• Account: username, email, firstname, lastname, biography
• Media: avatar, avatar_small
• Role: user_role (1=admin, 2=editor, 3=author, 4=subscriber)
• Settings: language, account_suspended, email_alerts, device_alerts
• Social links: link_website, link_facebook, link_x, link_instagram, link_youtube, link_twitch, link_linkedin
• Filters: user_role, account_suspended, language, search

Media API

The most feature-rich resource — manage your full media library including remote imports, metadata editing, and WebP cache generation:

• Core: file_name, file_path, file_type (MIME), file_size, resolution
• Metadata: title, credit, description
• Cache: cache_versions (JSON with small/medium/large WebP variants)
• Special actions:
  • upload_from_url — Import an image directly from a remote URL
  • update_metadata — Update title, credit, and description fields
  • create_all_cache — Generate resized and WebP-optimized variants
  • delete_all_cache — Remove all cached size variants
• Filters: file_type, search, date_from, date_to

Common Request Parameters

All resources support these standard parameters for pagination, field selection, and sorting:

• fields — Array of specific fields to return (reduces response size)
• filters — Object of key/value filter pairs
• limit — Number of results per page
• offset — Pagination offset
• order_by — Sort field and direction

Response format is consistent across all resources: {"status": "success", "data": {...}} on success, or {"status": "error", "message": "..."} on failure.

A Quick Example: Publishing an Article via API

Here's what a minimal API call to publish a new article looks like:

POST https://yoursite.newspro.io/functions/api
Authorization: Bearer YOUR_SECRET_KEY

{ "resource": "article", "action": "add", "title": "My First API Article", "body": "<p>Hello from the API!</p>", "category_id": 1, "status": 1, "approved": 1 }

That's it. One request. Article published.

Why llms.txt Makes the API Even More Powerful

The combination of a clean REST API and llms.txt documentation creates something genuinely new: an AI-native publishing platform. Paste newspro.io/llms-full.txt into any AI assistant — Claude, ChatGPT, Gemini — and ask it to write working integration code on the spot.

No digging through docs. No trial and error. Just describe what you want to build and let the AI write the API calls. Here are practical prompts that now work accurately because of NewsPro's llms.txt:

• "Show me how to bulk-import 50 articles into NewsPro from a CSV file using the API"
• "Write a Python script that creates a new category and publishes 3 articles under it"
• "How do I suspend a user account and transfer their articles to another author via the API?"
• "Generate a Make.com scenario that posts to NewsPro whenever a new item appears in an RSS feed"

How to Get Started in 3 Steps

1. Get your API token — Log into your NewsPro dashboard, go to Settings → API, and copy your secret key.
2. Make your first request — Send a POST to https://yoursite.newspro.io/functions/api with your Bearer token and a JSON body containing resource and action.
3. Use AI to accelerate — Paste newspro.io/llms-full.txt into any AI assistant and describe the integration you want to build.

Full API Documentation

The complete reference is available in both human-readable Markdown and AI-optimized formats:

• API Overview — Authentication, request format, response structure
• Articles API — Create, read, update, delete articles
• Categories API — Manage category hierarchy and navigation
• Pages API — Static page management
• Breaking News API — Urgent alerts and ticker management
• Users API — Full account and role management
• Media API — File library, uploads, metadata, and cache
• llms.txt — Platform summary for AI assistants
• llms-full.txt — Complete API reference for AI assistants