mcpField
S
sjinn-image-generation

sjinn-image-generation

github.com/sjinn-aiCreative & media075 installsUpdated 2w ago
Repository

About sjinn-image-generation

sjinn-image-generation is a skill for using the sjinn CLI to generate or transform images, including with reference images. It guides image-model selection, asynchronous tasks, and result handling.

Key features

  1. 01Text-to-image generation
  2. 02Image variation from URLs or files
  3. 03Multiple reference images
  4. 04Asynchronous task submission
  5. 05JSON output and result handling

Use cases

  1. 01Create images from text prompts
  2. 02Transform an existing image
  3. 03Generate from multiple visual references
  4. 04Save a generated result locally

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/sjinn-ai/skills --skill sjinn-image-generation

After installing, type/sjinn-image-generationin Claude Code to confirm it loaded.

Bundled filesreferences/models.md

Details

From the project README

SJinn Image Generation

Use this skill when the user wants to create an image, transform an image, use a reference image, or download an image result through SJinn.

Prerequisites

Before generating, quickly verify the CLI and login:

sjinn --version
sjinn auth whoami

The CLI must be 0.1.19 or newer for GPT Image 2.5 Flare and Sunburst. If either command fails or the CLI version is older, use sjinn-setup first.

Workflow

  1. Choose a model and options. For model-specific options, see references/models.md. GPT Image 2.5 models have different aspect ratios from gpt-image-2; do not reuse its auto, 3:2, or 2:3 values.
  2. Prefer JSON output so results are easy to parse.
  3. Use --async when the user wants a task ID immediately.
  4. Use --download [path] only when the user asks to save the result locally.

Commands

Text to image:

sjinn image generate --prompt "a clean product render of a glass lamp" --model gpt-image-2 --aspect 1:1 --json

Image variation from a URL:

sjinn image generate --prompt "make this product photo more cinematic" --image "https://example.com/input.png" --model nano-banana --json

Image variation from a local file:

sjinn image generate --prompt "turn this sketch into a polished concept render" --image "./sketch.png" --model seedream-v4 --json

GPT Image 2.5 Flare text to image:

sjinn image generate --prompt "a clean product render of a glass lamp" --model gpt-image-2.5-flare --aspect 3:4 --resolution 2K --json

GPT Image 2.5 Sunburst with multiple reference images:

sjinn image generate --prompt "keep the product from the first image and use the lighting from the second" --model gpt-image-2.5-sunburst --image "./product.png,./lighting.webp" --aspect 16:9 --resolution 4K --json

Both GPT Image 2.5 models support up to 16 reference images and default to --aspect 1:1 --resolution 1K when those options are omitted.

Async:

sjinn image generate --prompt "a futuristic workspace at sunrise" --model nano-banana-2 --resolution 4K --async --json

Output Handling

On success, return the task ID, primary URL, any additional URLs, model, and elapsed time. If status is created, tell the user to query the task with sjinn-task-status.

On error, summarize the CLI error plainly and suggest sjinn-setup only when the error indicates the CLI is missing or unauthenticated.

Comments

Markdown supported