A skill that makes Claude work like the art director on a product shoot. You give it photos of your product and say what the pictures are for. It asks the two or three questions that change the shoot, plans the set shot by shot, and writes finished prompts for the image tool you use, whether that is Nano Banana Pro in Google AI Studio, GPT Image in ChatGPT, Midjourney or FLUX. When you paste the results back, it checks each one against a realism checklist and writes the prompt that fixes what failed.
Inside are the rules the Tastebud studio shoots by. The product is reproduced exactly, down to the label, the shape and every word printed on it. Liquids, glass, metal and fabric behave the way they do in front of a real camera, the light has a direction, and there is a real shadow where the product meets the surface. There are playbooks for twenty-four categories, from skincare and fragrance to footwear, coffee, jewellery and electronics, and Claude reads the one your product belongs to before it plans anything.
Claude cannot take the photographs itself in most places, and the skill does not pretend otherwise. What it gives you is the thinking a good art director brings to a shoot, written as prompts you can paste. The check afterwards is what saves a set: most AI product pictures fail on the label or the shadow, and a second, specific prompt usually fixes them.
Claude skill · version 1.0
Product photo shoot
Checking your access.
In the download · 26 files
- SKILL.mdThe art director's method: the realism bar, what to ask, the six angles, the job of each frame, prompts for each image tool, and the review.
- README.mdHow to install it in claude.ai or Claude Code, how to start, and the licence.
- industries/alcohol.mdWine, spirits, beer and ready-to-drink cocktails: backlit glass, cold and warm light, and drinks that behave like drinks.
- industries/apparel.mdClothes off the body: flat lay, ghost mannequin, hung, folded and detail shots, with fabric and drape.
- industries/bags-leather.mdBags, wallets, belts and luggage: silhouette, grain, stitching and hardware.
- industries/bath-body.mdBody care, soap and scrubs: wet hero shots, lather and lotion texture.
Install it and begin
01Install it in Claude CodeTerminal mkdir -p ~/.claude/skillsunzip ~/Downloads/product-photo-shoot.zip -d ~/.claude/skills/# Start it with /product-photo-shoot. If this is your first skill, type /reload-skills once.On claude.ai there is nothing to unzip: upload the zip as it is. Every step is on Install and use a Claude skill.
02Your first message to ClaudeText Use the product photo shoot skill. I have attached photos of [PRODUCT], shot on my phone. The pictures are for [WHERE THEY WILL RUN]. Our brand looks [BRAND FEEL IN A FEW WORDS]. I make images with [YOUR IMAGE TOOL]. Plan the set first, then give me the prompts one shot at a time.Attach the clearest photos you have, front and back. The label is what the skill protects first.
Worked example
Adding it to claude.ai
- 01Download the zip from this page and leave it zipped.claude.ai takes the zip itself, with the skill's folder inside it.
- 02In claude.ai, open Customize, then Skills. Press the + button, choose Create skill, then Upload a skill.Anthropic's help centre said skills work on every plan, the free one included, when we checked on 28 September 2026.
- 03Select the product-photo-shoot zip you downloaded.The skill appears in your list. Check that its toggle is on.
- 04Check that code execution is on in your settings.Anthropic's help centre lists it as a requirement for skills.
- 05Start a new chat, attach your product photos and paste the first message below.Claude reads the skill when your request matches what it is for.
What to do it with
- ClaudeFree tier, paid above it
A language model that writes, edits, reads long documents and runs tools, including agents that work inside a codebase.
Bad at It is agreeable by default, so it will improve a bad idea instead of telling you it is bad. Ask it to argue against you and the quality of the answer changes.
- Claude CodeSubscription
An agent that reads a whole repository, edits files, runs the tests and explains what it changed.
Bad at It is fast enough to make a large mistake quickly. Work on a branch, read the diff, and never let it push. Watch the starting mode: on Pro, Max and Team plans it opens in auto mode, which Anthropic's own list says may read a .env file and send the key in it to the API that key belongs to. Switch it off.
- Google AI StudioFree tier, paid above it
Where Gemini API keys, projects, billing and spend caps are made, with a playground for the models.
Bad at Free-tier prompts are used to improve Google's products, and the billing screens changed several times in 2026.
- GPT ImagePay per use
Generates and edits images with unusually good instruction-following, and can hold text inside a picture better than most.
Bad at Faces drift between generations unless you feed a reference every time. Skin can come out plasticky at default settings, and a heavy fidelity setting costs many times more and takes many times longer.
Why it matters
A prompt that says "make it look premium" gets the same glossy picture everyone else gets. This skill makes Claude ask what the product is made of, where the pictures will run and how the brand already looks, and then decide the light, the surface and the angle before anything is generated.
- Type
- technique
- Period
- ongoing