Automate Image Conversion with the Contenta CLI, MCP Server and AI Agents

Published July 2, 2026 · 9 min read

Most people meet Contenta Converter through its Windows wizard: drop in a folder of photos, pick a format, click convert. But the wizard is only one way to drive it. The exact same conversion engine — 60+ formats, RAW from 600+ cameras, resizing, effects — also ships with a full command-line interface, a built-in MCP server, and ready-made skills for AI coding agents.

This post walks through all three layers: scripting with the contenta CLI, wiring the MCP server into Claude Desktop or Claude Code, and letting an AI agent run your whole image pipeline from a plain-English request. Everything shown here ships in the standard installer — no separate download, no extra license.

The contenta CLI: the same engine, scriptable

The executable is called contenta and installs together with the desktop app — same installer, same license. You will find it in C:\Program Files\ContentaSoft\Contenta Converter PREMIUM\; add that folder to your PATH and it is available from any terminal.

It exposes 13 commands: convert, batch, info, effects, formats, pdf-album, slideshow, ai-transform, watch, profile, register, status and serve. Inputs are positional — you point a command at a file or folder and add options. Converting a Canon RAW file to a high-quality JPEG is one line:

contenta convert photo.cr2 --format jpg --quality 92

Batch processing works the same way — point it at a folder. This resizes an entire directory of product photos to fit within 2000 × 2000 pixels and saves them as JPEGs:

contenta batch ./photos --output ./ready --format jpg --resize 2000x2000 --resize-mode fit

Every command accepts --json for machine-readable output and returns proper exit codes, so the CLI drops cleanly into PowerShell scripts, bash pipelines, CI jobs and scheduled tasks.

Real automation recipes

Two patterns cover most of what our users automate. Both run unattended once set up.

1. The self-running watch folder

The watch command monitors a folder and converts every new or changed image automatically. Start it once and it keeps running — drop files in, get converted files out:

contenta watch ./incoming --output ./processed --format webp

Run it at logon via Task Scheduler and you have a zero-click conversion service: scans, phone dumps and camera imports land in the incoming folder, and web-ready WebP files appear in the output folder seconds later.

2. The e-commerce resize pipeline

Marketplaces are strict about image dimensions — Amazon, for example, wants product photos on a 2000 × 2000 canvas. One batch command turns a memory-card dump into an upload-ready folder:

contenta batch ./photos --output ./ready --format jpg --resize 2000x2000 --resize-mode fit

Wrap it in a two-line script and every new product shoot goes from camera to marketplace in a single command — RAW files included, since the engine decodes 600+ camera formats natively.

The MCP server: hand the engine to an AI agent

MCP (Model Context Protocol) is the open standard AI assistants use to call external tools. Contenta Converter ships with a built-in MCP server — you start it with a single command:

contenta serve

It exposes 10 tools — single and batch conversion, format detection, metadata reading and writing, AI background removal, PDF albums, slideshows and more — and it works during the free trial. To connect Claude Desktop, add this to your claude_desktop_config.json:

{
  "mcpServers": {
    "contenta": {
      "command": "C:\\Program Files\\ContentaSoft\\Contenta Converter PREMIUM\\contenta.exe",
      "args": ["serve"]
    }
  }
}

From then on you can ask your assistant things like “convert every HEIC in my Downloads folder to JPEG” and it calls the local engine directly. Your images never leave your machine — the agent orchestrates, the conversion runs locally.

Claude Code skills: image processing in plain English

If you use Claude Code, there is an even shorter path: the /contenta-image-processing skill teaches the agent the whole CLI — commands, options, platform size presets and common recipes. Install it by cloning the public repo and copying the skills into place:

git clone https://github.com/brunojanvier77/ContentaSoft
cp -r ContentaSoft/skills/* ~/.claude/skills/

Then just describe the job in your own words:

• “Resize every photo in this folder for Amazon and Etsy”
• “Convert these CR2 files to JPEG at 92% quality”

The skill translates the request into the right contenta commands, runs them, checks the exit codes and reports back. You review the plan; the agent does the typing.

Everything is documented on GitHub

The complete reference lives in a public GitHub repository: accurate CLI documentation for every command and option, the full MCP server reference with all 10 tools, and the installable Claude Code skills for Contenta Converter and the other ContentaSoft products.

github.com/brunojanvier77/ContentaSoft

The docs are maintained alongside the product, so what you read there matches the binary on your disk. Copy the recipes straight into your own scripts, or open an issue if you hit an edge case.

Frequently Asked Questions