ai-product-photography

Overview

What this skill does

ai-product-photography is a command-line driven AI product photography workflow built on top of the inference.sh CLI (infsh). It helps you generate professional, photorealistic product images on demand: clean studio packshots, lifestyle scenes, mockups, and marketing visuals for online stores and campaigns.

Instead of manually setting up prompts and model calls every time, this skill gives you a repeatable pattern for calling image models through infsh to produce:

• Studio shots on white or neutral backgrounds
• Lifestyle product images in context
• E-commerce packshots tuned for Amazon or Shopify
• Creative marketing and advertising visuals

Who it is for

This skill is designed for:

• E‑commerce and Amazon sellers who need consistent product photos at scale
• Shopify and DTC brands creating catalog and campaign imagery
• Marketing and social media teams producing fresh visuals for ads and posts
• Developers and technical marketers comfortable with the terminal and looking to automate product image generation

If you want to script or templatize product photos rather than using a point‑and‑click UI, ai-product-photography is a good fit.

Core capabilities

Backed by inference.sh, ai-product-photography can use multiple image models (as documented in the upstream guide):

• FLUX Dev (falai/flux-dev) – high‑quality, detailed product shots
• FLUX Schnell (falai/flux-schnell) – fast iterations while you refine prompts
• Imagen 3 (google/imagen-3) – realistic, natural‑looking photos
• Grok (xai/grok-imagine-image) – more creative, stylized variations
• Seedream – oriented toward product and commercial visuals (see upstream table for exact app ID)

Typical use cases include “product photography”, “product shot”, “amazon product photo”, “shopify images”, “studio product shot”, “lifestyle product image”, “packshot”, and “product mockup”.

When this skill is (and isn’t) a good fit

Use ai-product-photography when:

• You already use or are willing to install the inference.sh CLI
• You want reproducible product image generation from the command line
• You need to plug AI product photos into scripts, CI, or content pipelines

It may not be ideal if:

• You need a purely visual, drag‑and‑drop editor instead of CLI workflows
• You are not comfortable installing and authenticating CLI tools
• You require heavy image post‑processing or layout design (this skill focuses on generation, not editing)

How to Use

1. Prerequisites

To run ai-product-photography as intended, you need:

• A working terminal environment (macOS, Linux, or Windows with a compatible shell)
• The inference.sh CLI (infsh) installed
• An inference.sh account and credentials to log in via the CLI

Follow the official CLI installation instructions referenced in the upstream SKILL.md:

• https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md

Install infsh, then verify:

infsh --help

If the command prints help text, you are ready to proceed.

2. Install the ai-product-photography skill

If you are using a skills-enabled agent environment, you can add the skill repository with:

npx skills add https://github.com/inferen-sh/skills --skill ai-product-photography

This pulls in the shared inferen-sh/skills repository and registers the ai-product-photography skill, so your agent or tools can reference its configuration, prompts, and conventions.

Once installed, open the skill folder and inspect:

• SKILL.md – high‑level description and capability notes
• guides/photo/ai-product-photography – workflow details and examples (if present)

3. Log in to inference.sh

Before generating any images you must authenticate the CLI:

infsh login

Follow the login flow (for example, copying an auth URL into your browser and confirming access). After a successful login, your CLI can call the configured models.

4. Generate your first product shot

The upstream quick start shows a minimal product photography call using FLUX Dev. You can use it as a template:

infsh app run falai/flux-dev --input '{
  "prompt": "Professional product photo of wireless earbuds on white surface, soft studio lighting, commercial photography, high detail"
}'

This command:

• Calls the falai/flux-dev app
• Sends a JSON input payload with a product photography prompt
• Returns a generated product image (the exact output handling depends on your infsh configuration)

Adjust the prompt text to match your product, background, and style requirements.

5. Choose the right model for your task

The skill documentation lists several supported models. Typical selection patterns:

• Use FLUX Dev when you want maximum quality for final Amazon or Shopify listing images.
• Use FLUX Schnell during prompt exploration where speed matters more than small quality differences.
• Use Imagen 3 for photorealistic, natural‑light product scenes.
• Use Grok to explore creative or stylized variants for social media posts or ads.
• Use Seedream (per upstream app ID) for targeted product/commercial imagery.

Switching models is as simple as changing the app run target, for example:

infsh app run google/imagen-3 --input '{
  "prompt": "Lifestyle photo of a ceramic coffee mug on a wooden table near a laptop, morning light, realistic, e-commerce focus"
}'

6. Craft effective product photography prompts

ai-product-photography does not lock you into rigid templates; instead, it encourages rich, descriptive prompts tuned for e‑commerce. Consider including:

• Product type and attributes – “matte black stainless steel water bottle, 750ml, logo centered”
• Background and environment – “on white seamless background”, “on kitchen countertop”, “in gym setting”
• Lighting and mood – “soft studio lighting”, “high key, bright, no harsh shadows”
• Usage context – “held in a hand”, “next to laptop”, “on bathroom shelf”
• Quality hints – “commercial product photography, high resolution, high detail”

Example studio shot:

infsh app run falai/flux-dev --input '{
  "prompt": "Studio packshot of a matte black stainless steel water bottle on pure white background, soft diffused lighting, sharp focus, commercial product photography, e-commerce listing image"
}'

Example lifestyle image:

infsh app run falai/flux-schnell --input '{
  "prompt": "Lifestyle photo of running shoes on a park path at sunrise, shallow depth of field, athlete in background out of focus, modern ad style, social media ready"
}'

7. Integrate with e‑commerce and marketing workflows

Because ai-product-photography is CLI‑based, it fits easily into scripts and pipelines you may already use for Amazon, Shopify, or social media:

• Batch generation scripts – loop over a list of SKUs and generate one or more images per product
• Content pipelines – incorporate infsh calls into build scripts that generate assets before publishing
• A/B testing – quickly create visual variations for ads and measure performance

For example, a simple Bash loop (conceptual example only):

#!/usr/bin/env bash

PRODUCTS=("wireless earbuds" "smartwatch" "fitness tracker")

for PRODUCT in "${PRODUCTS[@]}"; do
  infsh app run falai/flux-dev --input "{
    \"prompt\": \"Professional studio photo of ${PRODUCT} on white background, soft studio lighting, e-commerce packshot, high detail\"
  }"
 done

Adjust paths and output handling according to your environment and the inference.sh documentation.

8. Explore the repository for advanced usage

To go beyond the quick start, open the skill’s source in the inferen-sh/skills repository:

• Review SKILL.md for the canonical description, tooling constraints (Bash / infsh), and triggers.
• Check under guides/photo/ai-product-photography for deeper photo workflows, style suggestions, or prompt recipes (if present in your version of the repo).

From there you can adapt:

• Standard prompt structures for your own brand guidelines
• Model choices for different product categories
• Shell scripts to automate repeated tasks

FAQ

Is ai-product-photography a standalone app or a CLI workflow?

ai-product-photography is a CLI‑centric skill, not a standalone GUI application. It relies on the inference.sh CLI (infsh) and the inferen-sh/skills repository. You run commands in your terminal (or via scripts) to generate product images.

What tools does ai-product-photography require?

According to SKILL.md, the allowed tool is:

• *Bash (infsh ) – meaning the workflow is designed to run through Bash scripts or shell commands that call the infsh CLI. You should have Bash (or a compatible shell) and the inference.sh CLI installed.

Which models can I use with this skill?

The upstream documentation lists several example models:

• falai/flux-dev (FLUX Dev)
• falai/flux-schnell (FLUX Schnell)
• google/imagen-3 (Imagen 3)
• xai/grok-imagine-image (Grok)
• Seedream (see the repository for the complete app ID)

You can switch among these by changing the app run target in your infsh command.

Can I use ai-product-photography for Amazon and Shopify listings?

Yes. The skill description specifically calls out e‑commerce, Amazon listings, and Shopify as core use cases. By combining studio‑style prompts (white background, clear lighting, sharp focus) with high‑quality models like FLUX Dev or Imagen 3, you can generate images suited for product detail pages.

Always confirm that generated images meet marketplace guidelines and your own brand standards.

Does this skill handle post‑processing or only generation?

ai-product-photography focuses on image generation through inference.sh models. It does not provide built‑in post‑processing (cropping, background removal beyond what the model renders, text overlays, etc.). For advanced editing, use image editing tools after generation.

Do I need prior experience with AI image models to use this?

You do not need deep model knowledge, but basic CLI familiarity is important. The skill’s quick start commands give you working examples; from there you mainly adjust prompts and choose the appropriate model for your needs.

How do I troubleshoot if a command fails?

If an infsh command errors:

1. Confirm that the CLI is installed and available: infsh --help.
2. Run infsh login again if you suspect an authentication issue.
3. Check that the app ID (for example, falai/flux-dev or google/imagen-3) matches values in the repository table.
4. Validate your JSON in the --input payload (matching quotes and braces).

If problems persist, compare your commands with the examples in SKILL.md and the related guide paths in the inferen-sh/skills repository.

Where can I see all the details of the ai-product-photography configuration?

Open the upstream repository path:

• https://github.com/inferen-sh/skills/tree/main/guides/photo/ai-product-photography

From there you can inspect SKILL.md and any additional guides, rules, or helper files. In the Agent Skills Finder interface, use the Files tab to browse the full tree and see exactly how the skill is defined.