product-photography Skill (inference.sh): AI Product Photos and Packshots

Makes Claude plan product photos (hero, packshot, lifestyle, scale, detail) and generate them with the inference.sh belt CLI.

Installs: 59.1K; on skills.sh
Stars: 17; on GitHub
Updated: Aug 2026; last repo commit

This skill teaches Claude product photography terms and shows belt app run commands that generate the images. It covers six shot types, camera angles, lighting, shadows, backgrounds, composition, an e-commerce image set and an edit-and-upscale workflow.

It is one SKILL.md. It needs the belt CLI from inference.sh and an inference.sh login, because the shots are generated by apps such as bytedance/seedream-4-5 and falai/flux-dev-lora.

Good for

Online sellers who need packshots and listing images
Small brands making hero and lifestyle shots without a studio
Designers who want prompt keywords for angles, lighting and shadows

Install product-photography

Claude Code

Copy this prompt.

Install the product-photography skill in Claude Code. Run this command:

npx skills add inference-sh/skills@product-photography

Then check that it loaded and tell me how to use it.

2
Paste it into Claude Code. It shows the change and asks before running it.

Requires Node.js.

The skill only does its image, video and search steps when the belt CLI is installed and logged in (SKILL.md Quick Start). The Homebrew and installer commands come from cli-install.md, which also lists npx and Scoop. The vendor README documents a Claude Code plugin route (/plugin marketplace add inference-sh/skills), but the repo has no .claude-plugin/marketplace.json (HTTP 404 on Nov 2, 2026) and the plugin would add every skill in plugin.json, so it is not listed here. This page installs from inference-sh/skills, the vendor's own repo.

From its listing on skills.sh. Source.

What's in the skill

When it loads

AI product photography with studio lighting, lifestyle shots, and packshot conventions. Covers angles, backgrounds, shadow types, hero shots, and e-commerce image requirements. Use for: product photos, e-commerce images, Amazon listings, packshots, lifestyle photography. Triggers: product photography, product photo, packshot, e-commerce photography, product shot, product image, studio photography, lifestyle product, amazon product photo, product listing image, hero shot, product mockup, commercial photography

Sections in SKILL.md

Quick Start
Shot Types
1. Hero Shot (Primary Image)
2. Packshot (E-Commerce White Background)
3. Lifestyle Shot
4. Scale Shot
5. Detail / Close-Up Shot
6. Group / Collection Shot
Camera Angles
Lighting Setups
Shadow Types
Background Guide
Composition Rules
E-Commerce Image Set
Product Categories
Food & Beverage

Key facts

Author: inference.sh
License: No LICENSE file in the repo; the README badge and plugin.json both say MIT
Installs on skills.sh: 11 (Nov 2, 2026)
GitHub stars: 797 (Nov 2, 2026)
Last commit: Oct 15, 2026 (repo)
What installs: 1 file: SKILL.md
Needs: The belt CLI from inference.sh and a belt login

What it includes

Logs in with belt and generates a clean studio packshot.
Runs code · Quick Start
Hero, packshot, lifestyle, scale, detail and group shots, each with a belt command.
Runs code · Six shot types
Five angles with the prompt keywords that produce them.
Camera angles
Five lighting looks with prompt keywords, from soft box to hard directional.
Lighting setups
Five shadow styles from no shadow to reflection, with when to use each.
Shadow types
Six background choices with prompt keywords.
Background guide
Six rules such as rule of thirds, negative space and odd numbers.
Composition rules
Lists a nine-image order for a listing and gives belt commands for four of them.
Runs code · E-commerce image set
Advice for food, jewelry, electronics, cosmetics and apparel products.
Product categories
Generates a base image, edits the background, then upscales for print.
Runs code · Image editing workflow
Eight common mistakes with the problem and the fix for each.
Common mistakes

From the official docs

Permissions

Medium risk

Claude can run belt commands that call AI apps through your inference.sh login. Read the first command before it runs and check app pricing before bigger jobs.

What it can do

Read your data
Create and change things
Delete things
Read-only mode

Watch out for

Claude runs belt commands for you
SKILL.md lists `Bash(belt *)` in allowed-tools and its commands call apps such as bytedance/seedream-4-5, falai/flux-dev-lora, bytedance/seededit-3-0-i2i and falai/topaz-image-upscaler. Each command's --input is passed to the named app, so read prompts before they run. github.com
Running apps can use credits
The belt CLI skill documents belt balance (check credits) and belt app pricing (see cost formula before running). Check the price before large or repeated runs. github.com
You install the belt CLI separately
cli-install.md offers a curl installer, Homebrew, npx and Scoop. It says the installer downloads a binary from dist.inference.sh, verifies its SHA-256 checksum, and uses no elevated permissions, background processes or telemetry. github.com

Security advisories

No published security advisories as of Nov 2, 2026.

Technical details

Talks to	inference.sh (belt app run). cli.inference.sh and dist.inference.sh (belt installer)
Sign-in and keys	SKILL.md tells you to run belt login. The files we read do not say where belt keeps the login.
Scopes	Bash(belt *) in SKILL.md allowed-tools

Questions

What does product-photography do?

It makes Claude plan and generate product images: hero shots, white-background packshots, lifestyle, scale and detail shots. It supplies prompt keywords for angle, lighting, shadow and background, and the belt commands that run the image apps.

How do I install product-photography?

Use the install box above. It installs the skill folder, and you also need the belt CLI, which the SKILL.md links to in cli-install.md. Then run belt login, as SKILL.md's Quick Start does.

Does it need an account or API key?

It needs the belt CLI and a belt login to inference.sh. SKILL.md names these apps: bytedance/seedream-4-5, falai/flux-dev-lora, bytedance/seededit-3-0-i2i and falai/topaz-image-upscaler. The belt CLI skill documents belt balance to check credits and belt app pricing to see the cost formula before running an app, so check pricing there.

Is it safe to use?

The skill's frontmatter allows Bash(belt *), so Claude can run belt commands that call AI apps on inference.sh. The files in this folder are one SKILL.md. Our advice: read the first belt app run command Claude proposes, and check an app's price with belt app pricing before large runs.

Is there another copy of product-photography?

Yes. skills.sh also lists a copy under 101-skills/superpowers, with many more installs. This page installs from inference-sh/skills, the vendor's own repo, and the facts here come from there. Pick one source so you do not get the skill twice.

What else does SKILL.md point to?

Its Related Skills section lists these installs from inference-sh/skills: ai-image-generation, flux-image and prompt-engineering. They are optional hand-offs, not requirements for the layout and spec advice in this skill.

Information

updated Nov 2, 2026

Repository: inference-sh/skills
Folder: guides/photo/product-photography

inference-sh/skills skills.sh

Sources

Docs, issue trackers and discussions this page is built from.

Show all 5

cli-install.md · github.com
product-photography SKILL.md · github.com
plugin.json · github.com
inference-sh/skills README · github.com
belt CLI SKILL.md · github.com