Flux Image

Generate images with FLUX models (Black Forest Labs) via inference.sh CLI. Models: FLUX Dev LoRA, FLUX.2 Klein LoRA with custom style adaptation. Capabilitie...

Security Scan

The skill's purpose (run FLUX models via the inference.sh CLI) is coherent, but it directs users to download-and-execute a remote install script and relies on a CLI login flow that isn't declared in the metadata — that install/login pattern is a moderate risk and should be reviewed before use.

Purpose & Capability

Name/description match the runtime instructions: the SKILL.md shows how to install and use the inference.sh CLI to run FLUX image models. No unrelated binaries, env vars, or config paths are requested.

Instruction Scope

Instructions are narrowly scoped to installing the inference.sh CLI and invoking infsh app run/sample commands; they do not ask the agent to read arbitrary system files or unrelated credentials. However, the doc tells the user to run 'infsh login' (an interactive auth step) and to pipe a remote install script to sh, which are broader actions than what the metadata declares.

Install Mechanism

There is no formal install spec in the registry metadata, but SKILL.md explicitly recommends running 'curl -fsSL https://cli.inference.sh | sh' (download-and-execute from a service domain). While the doc claims checksum verification is available, piping a remote script to sh (and relying on a project-hosted binary distribution) is higher risk than using a reviewed package manager. Verify checksums and inspect the install script before running.

Credentials

The skill declares no required environment variables or credentials, which aligns with the metadata. But the instructions include 'infsh login' (implying account credentials or tokens will be used/stored). The skill does not declare where such credentials will be stored or what scope they have — the login step is not reflected in the declared requirements.

Persistence & Privilege

The skill is instruction-only and not marked always:true. It does not request elevated or persistent platform-wide privileges in the metadata. Installing the CLI would create a local binary (normal for a CLI) but that is a user action, not an automatic persistent modification by the skill.

What to consider before installing

This skill appears to do what it says (drive the inference.sh CLI to run FLUX models), but pay attention to the install and auth steps before proceeding. Don't blindly run 'curl https://cli.inference.sh | sh' — download and inspect the script first and verify the SHA-256 checksums linked in the SKILL.md. Expect to provide/login with an inference.sh account or token; confirm where the CLI stores that token and whether you trust the service. If you prefer lower risk, install the CLI from a package manager (if available) or run it in a sandboxed environment or container.

FLUX Image Generation

Generate images with FLUX models via inference.sh CLI.

Quick Start

curl -fsSL https://cli.inference.sh | sh && infsh login

infsh app run falai/flux-dev-lora --input '{"prompt": "a futuristic city at night"}'

Install note: The install script only detects your OS/architecture, downloads the matching binary from dist.inference.sh, and verifies its SHA-256 checksum. No elevated permissions or background processes. Manual install & verification available.

FLUX Models

Model	App ID	Speed	Quality	Use Case
FLUX Dev LoRA	falai/flux-dev-lora	Medium	Highest	Production, custom styles
FLUX.2 Klein LoRA	falai/flux-2-klein-lora	Fastest	Good	Fast iteration, 4B/9B sizes

Examples

High-Quality Generation

infsh app run falai/flux-dev-lora --input '{
  "prompt": "professional product photo of headphones, studio lighting, white background"
}'

Fast Generation (Klein)

infsh app run falai/flux-2-klein-lora --input '{"prompt": "abstract art, colorful"}'

With LoRA Custom Styles

infsh app sample falai/flux-dev-lora --save input.json

# Edit to add lora_url for custom style
infsh app run falai/flux-dev-lora --input input.json

Image-to-Image

infsh app run falai/flux-dev-lora --input '{
  "prompt": "transform to watercolor style",
  "image_url": "https://your-image.jpg"
}'

For Other Image Tasks

# Image editing with natural language
infsh app run falai/reve --input '{"prompt": "change background to beach"}'

# Upscaling
infsh app run falai/topaz-image-upscaler --input '{"image_url": "https://..."}'

Related Skills

# Full platform skill (all 150+ apps)
npx skills add inference-sh/skills@inference-sh

# All image generation models
npx skills add inference-sh/skills@ai-image-generation

# Upscaling
npx skills add inference-sh/skills@image-upscaling

Browse all apps: infsh app list

Documentation

• Running Apps - How to run apps via CLI
• Image Generation Example - Complete image generation guide
• Streaming Results - Real-time progress updates