flux-kontext

Edit a single source image with precise localized changes while preserving subject identity.

31|9|Updated Apr 30, 2026
One-click install
npx skills add https://github.com/prime-skills/runcomfy-agent-skills --skill flux-kontext-prime-skills
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: flux-kontext
Source: https://github.com/prime-skills/runcomfy-agent-skills/tree/main/flux-kontext
Command: npx skills add https://github.com/prime-skills/runcomfy-agent-skills --skill flux-kontext-prime-skills

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve?

This Skill helps you make precise, localized image edits without losing the subject’s identity, pose, composition, or visual details.

Core Features & Use Cases

  • Targeted Image Editing: Add, remove, or replace a specific element in a single source image while preserving everything else.
  • High-Fidelity Preservation: Use preservation-led prompts for reliable edits to people, products, branding, backgrounds, and visual details.
  • Model Routing Guidance: Choose Flux Kontext for single-image edits, or route batch edits, multilingual text editing, and image generation to more suitable sibling models.
  • Use Case: Update a product photo by replacing the label text or add an accessory to a portrait while keeping the original lighting, pose, and identity intact.

Quick Start

Ask the Flux Kontext skill to edit a publicly accessible source image with one precise instruction, such as adding an orange umbrella while preserving the subject’s face, pose, clothing, and background.

Frequently Asked Questions about flux-kontext

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I make precise localized image edits without losing the subject's identity?

To perform precise image editing, submit a declarative prompt to the Flux 1 Kontext Pro edit endpoint. This approach ensures high-fidelity preservation of the subject's identity, pose, and lighting during targeted photo retouching or product imagery updates.

Can I edit a product photo and replace label text while keeping the original lighting intact?

Editing a product photo to replace label text while keeping original lighting intact is supported by Flux Kontext. You provide a public HTTPS image URL and a declarative prompt to update brand assets without altering the surrounding visual details.

Do I need a public HTTPS image URL to use Flux Kontext for photo retouching?

Yes, a public HTTPS image URL is required for photo retouching with Flux Kontext. The RunComfy CLI processes your declarative prompt against this accessible source image to apply high-fidelity localized changes.

What is the best way to add an accessory to a portrait while preserving clothing and background?

Adding an accessory to a portrait while preserving clothing and background is best handled through single-image editing with Flux Kontext. It applies targeted changes via declarative prompts without altering the subject's pose or identity.

Are there limitations when using Flux Kontext for batch edits or multilingual text editing?

Flux Kontext is limited to single-image editing scenarios. Batch edits, multilingual text editing, and general image generation should be routed to more suitable sibling models to ensure reliable results.