AI Agents

Machine-readable documentation for LLMs, AI agents, and automation tools.

llms.txt

We publish llms.txt files — a standard for making websites readable by large language models:

  • llms.txt — structured overview for LLM context windows
  • llms-full.txt — complete API reference in plain text
Feed llms-full.txt to your LLM for complete API coverage in a single context load.
bash
# Fetch the full reference
curl https://img.pro/llms-full.txt

OpenAPI Spec

Full OpenAPI 3.1 spec → /api/openapi.

MCP

img.pro works as a tool in any Model Context Protocol (MCP) setup. Point your agent at the OpenAPI spec or llms-full.txt and generate the full tool set — upload, list, get, transform, and delete — straight from the spec.

The example below is a minimal tool that imports an image from a URL and returns a transformable img.pro URL. For the full tool set — create, list, get, transform, delete — generate it from the OpenAPI spec with any OpenAPI code generator.

json
{
  "name": "img_pro_import",
  "description": "Import an image from a URL into img.pro and get back a transformable URL",
  "parameters": {
    "url": { "type": "string", "description": "Image URL to import" }
  },
  "endpoint": "POST https://api.img.pro/v1/images",
  "headers": { "Authorization": "Bearer YOUR_API_KEY" }
}
Every response's url is directly transformable — an agent can resize or convert by appending query params (e.g. ?size=m, ?format=webp) with no extra API call. See Transforms.