For photographers extending a real shoot, and AI image creators who need consistent characters, products and sets. It diagnoses a reference frame, finds the coverage gaps, and writes 9-grid and single-shot prompts locked to the same world, light and lens.
By Fia Houston-Hamilton · 2 resources
One good image is not a campaign. The moment you ask for a second angle, AI drifts: the light moves, the wardrobe changes, props appear that were never on set.
Pocket DoP works the way a Director of Photography does. It designs coverage, not concepts.
Coverage is where you start. Then it keeps going.
Ask for a new scene, a new moment or a whole campaign and it builds them in the same visual language, no new reference frame needed. Ask for a storyboard, a sequence, an opening or a transition and it designs the shot flow in order. Ask for video and it thinks in movement: subject action, camera action, how the two interact, and where the cut lands.
One frame becomes a world.
This is the workflow I recommend, and it removes the most common friction point.
Run Pocket DoP inside ChatGPT through the Eden MCP and you generate in the same conversation. It writes the prompt, the model makes the image, and the next shot carries on in the same chat with the world lock intact. No copy-paste, no losing your reference between tools.
Connect Higgsfield to ChatGPT or Claude and you can generate image and video the same way, from the same locked world.
Some people want the whole world laid out in front of them. Take the prompts Pocket DoP writes and build in a space like Firefly Boards, where you can arrange every frame and push the creative freedom further by hand.
It is not a single-click workflow, but it is a powerful one. There is no right way to run it, only the way that fits how you work.
No new props, no new locations, no change of light, time of day or wardrobe, no invented story, and no switching your aspect ratio unless you ask. It extends what exists.
Photographers who want the shots they didn't get on the day. Brands and creators who need the same character, product or set to hold across a whole campaign in AI image or video.
Pocket DoP lives in Eden on web and phone, and it works inside ChatGPT or Claude through the Eden MCP. If you already pay for ChatGPT or Claude, use your own plan with no second subscription.
Buy once, then open Pocket DoP in the AI app you already use. The instructions and knowledge come with it.
Pocket DoP as a Custom AI in your own Eden account. Send it one reference frame and it walks you from diagnosis to finished coverage prompts.
Any single frame with a clear scene: people, products, interiors, landscapes. It works only from what is visible, so it will never suggest a shot the frame cannot support.
Yes. Give it a frame from your shoot and it builds the angles you didn't capture, locked to the same light and set.
No. It asks you to choose between options at each stage and writes every prompt itself.
A free Eden account comes with 50 credits, and each message uses one. If you already pay for ChatGPT or Claude, connect the Eden MCP and run Pocket DoP on your own plan instead. Eden credits never expire, so you can buy a pack and use it at your own pace. No monthly subscription needed.
It opens in your Eden account straight away, on web and phone. You keep it, and when Fia improves it, your copy updates automatically.
Fia Houston-Hamilton