imagic Max · built-in MCP server

Score every frame before you look at any of them

Automatic quality scoring for sharpness, exposure, closed eyes and composition across a whole shoot, readable by your AI agent through MCP.

What you type

Score this shoot and show me the twenty worst frames and why they scored badly.

Scoring is the step that makes everything after it cheap. Once every frame carries a number and a reason, culling, ranking and re-selection are all just queries.

The scores are readable by the agent, so a shoot can be interrogated in plain language rather than clicked through.

What happens, step by step

1

Run the pass

analyze_photos scores pending photos for sharpness, exposure, eyes and composition.

2

Ask questions

list_photos and get_photo expose scores, EXIF and edit state, so the agent can explain why a frame was dropped.

3

Check the shape of the shoot

get_library_stats returns counts per status, which is how you know where you are.

Why it is worth doing this way

A score with a reason attached is auditable. You are not asked to trust a black box; you can ask why any individual frame landed where it did.

The tools behind it

Nothing here is a black box. This workflow is these tools, in this order, and you can read exactly what each one does:

analyze_photos list_photos get_photo get_library_stats

Getting set up

You need imagic Max and an MCP client. The free 7-day trial includes Max with no card, and connecting a client takes about three minutes:

Claude Code Claude Desktop Cursor VS Code and GitHub Copilot every client

Questions

What does it measure?

Technical quality such as sharpness and exposure, subject issues such as closed eyes, and composition, combined into a score per frame.

How long does the pass take?

Roughly 30 to 90 seconds per 1,000 photos on a modern machine.

Other things to ask for

Try it on your own shoot

Free for 7 days, everything included, no card. Then imagic Max is €99 once, with no subscription.