How do I install the Image Edit skill?

Run npx skillmds@latest add yigityildiz0/image-edit in your terminal (requires Node.js), paste this page's agent-chat prompt into Claude, Cursor, or any MCP-connected agent, or download the SKILL.md file and copy it into your agent's skills directory.

Is Image Edit safe to use?

This skill has not completed SkillMD's automated safety review yet. SkillMD never runs a skill's scripts for you; review the SKILL.md before installing.

Which AI agents work with Image Edit?

This skill is tagged as working with Claude Code, Claude.ai, OpenAI Codex. SKILL.md is an open format, so most agents that read a skills directory can load it too.

Is Image Edit free to use?

Yes. Installing skills from SkillMD is free. This skill is licensed under MIT.

Who published Image Edit?

yigityildiz0 (@yigityildiz0) published this skill. Their other Agent Skills are listed on their SkillMD profile.

Image Edit by yigityildiz0 — Agent Skill

Name: Image Edit
Author: yigityildiz0

Image Edit

Plan, execute, and verify image edits with the active host's already available local or approved image-editing capability. Use for object/background changes, cleanup, inpainting, compositing, relighting, text replacement, resizing, or style changes; preserve the source and never switch providers/models or upload images automatically.

yigityildiz0 2d9c08d 2.3 KB Updated Sep 17, 2026

File contents

Portable Image Editing

1. Inspect

View the source at adequate resolution. Record dimensions, color profile, alpha, orientation, compression, visible text, protected identity/features, and likely privacy/copyright concerns. Keep the original untouched.

2. Define the edit contract

State:

what must change;
what must remain identical;
edit region or mask;
output size/aspect/format/transparency;
realism/style and lighting perspective;
text spelling/language if applicable;
acceptance criteria and allowed variation.

For people, explicitly preserve identity, anatomy, skin texture, age cues, pose, and expression unless the request changes them. Avoid inferring or altering sensitive traits.

3. Choose an available path

Use an image-edit tool already exposed by the active host, an installed local workflow, or a provider the user already authorized. Do not choose a provider/model, spend credits, install a CLI, or upload a private image merely because a reusable skill names one. If no suitable capability exists, produce the edit plan/prompt and clearly state the missing tool.

Use masks for localized changes. Prefer one controlled edit over broad regeneration when preservation matters. For text replacement, specify exact text, location, alignment, type character, and unchanged surroundings.

4. Iterate conservatively

Change one major factor per iteration. Keep source, mask, prompt/parameters, seed when available, and outputs. Do not repeatedly regenerate after the acceptance criteria are met.

5. Verify

Compare source and output at full image and 100% crops. Check protected regions, seams, edges, hands/faces, reflections, shadows, geometry, repeated textures, text accuracy, color profile, alpha, dimensions, and compression. Ensure removed content is not visibly retained when the request requires removal.

Report tool/path used, edit contract, output file, checks performed, deliberate deviations, and whether the image left the local machine.

---name: image-editdescription: Plan, execute, and verify image edits with the active host's already available local or approved image-editing capability. Use for object/background changes, cleanup, inpainting, compositing, relighting, text replacement, resizing, or style changes; preserve the source and never switch providers/models or upload images automatically.license: MIT---# Portable Image Editing## 1. InspectView the source at adequate resolution. Record dimensions, color profile, alpha, orientation, compression, visible text, protected identity/features, and likely privacy/copyright concerns. Keep the original untouched.## 2. Define the edit contractState:- what must change;- what must remain identical;- edit region or mask;- output size/aspect/format/transparency;- realism/style and lighting perspective;- text spelling/language if applicable;- acceptance criteria and allowed variation.For people, explicitly preserve identity, anatomy, skin texture, age cues, pose, and expression unless the request changes them. Avoid inferring or altering sensitive traits.## 3. Choose an available pathUse an image-edit tool already exposed by the active host, an installed local workflow, or a provider the user already authorized. Do not choose a provider/model, spend credits, install a CLI, or upload a private image merely because a reusable skill names one. If no suitable capability exists, produce the edit plan/prompt and clearly state the missing tool.Use masks for localized changes. Prefer one controlled edit over broad regeneration when preservation matters. For text replacement, specify exact text, location, alignment, type character, and unchanged surroundings.## 4. Iterate conservativelyChange one major factor per iteration. Keep source, mask, prompt/parameters, seed when available, and outputs. Do not repeatedly regenerate after the acceptance criteria are met.## 5. VerifyCompare source and output at full image and 100% crops. Check protected regions, seams, edges, hands/faces, reflections, shadows, geometry, repeated textures, text accuracy, color profile, alpha, dimensions, and compression. Ensure removed content is not visibly retained when the request requires removal.Report tool/path used, edit contract, output file, checks performed, deliberate deviations, and whether the image left the local machine.

Frequently asked questions