Fix logo

Fix

Organization
with-pebbly
fix

Change a sprite that already exists — the user's own work or your earlier output — without destroying what is already right. Use for edits, touch-ups, style corrections and "make it more X" requests.

Overview

Publisherwith-pebbly
Repositoryaseprite-ai-artist
Skill namefix
Stars
456
Forks
28
Bundled files
Instructions only
LicenseMIT
Links
  • Markdown instructions

    A SKILL.md file the model loads on demand, so it only costs tokens when a request actually matches.

  • Works with any LLM

    AI skills are plain Markdown, not provider-specific code, so this works with GPT, Claude, Gemini, Grok, or a local model.

  • Self-contained

    Everything the model needs lives in the instructions — no extra files to sync.

  • Open source

    Published by with-pebbly on GitHub. Read the source before you install it.

Installation

Install the Fix AI skill in TypingMind to use it with any LLM, or drop it into another agent that reads SKILL.md.

1

Install in TypingMind

TypingMind installs a skill straight from its GitHub folder — it reads SKILL.md, bundles the resource files, and stores the result locally.

  1. Open the app and go to Plugins → Skills.
  2. Choose "Install from GitHub".
  3. Paste the skill folder URL below and confirm.
    https://github.com/with-pebbly/aseprite-ai-artist/tree/main/skills/fix
  4. Enable the skill in any chat where you want it available.
Plugins → Skills → Add skill → From GitHub URL, then paste the folder URL and press Continue.
2

Install in another agent

Any agent that reads the Agent Skills format can use this skill — copy the folder into that agent's skills directory.

Claude Code — .claude/skills
git clone --depth 1 https://github.com/with-pebbly/aseprite-ai-artist.git /tmp/aseprite-ai-artist
mkdir -p .claude/skills
cp -r /tmp/aseprite-ai-artist/skills/fix .claude/skills/fix
Restart Claude Code after copying so it picks up the new skill.

Use it in TypingMind

Enable Fix in any TypingMind chat and the model takes it from there. Its name and description sit in the system prompt, and the moment a request matches, the model loads the full instructions itself — you never invoke it by hand, and it costs no tokens until it is actually used.

The model loads Fix on its own as soon as a request matches it.

Works with any AI model

AI skills are plain Markdown instructions rather than provider-specific code, so Fix is not tied to the model it was written for. Install it once in TypingMind and use it with GPT-5, Claude, Gemini, Grok, DeepSeek, Mistral, Llama, or a local model you run yourself — all on your own API keys.

  • Loaded only when it is needed

    The system prompt carries just the name and description. The instructions are fetched on the first matching request, so an idle skill costs nothing.

  • Switch models mid-chat

    Because the skill is instructions rather than code, changing model does not break it — the next model reads the same SKILL.md.

Skill instructions

This is the SKILL.md content the model loads. Read it before installing — a skill is instructions your model will follow.

Fix or iterate on existing art

Editing someone's work has a failure mode that drawing from scratch does not: breaking something that was already correct. The whole procedure is built around not doing that.

Procedure

1. Understand what is there before changing it

preflight
sprite_info
look op="preview"

For a targeted edit, also look op="ascii" on the region. You cannot edit pixels precisely from a description.

Never assume layer names, frame counts or the palette. Drawing into the wrong layer is the most common way an agent damages someone's file. When two groups hold same-named layers (left/arm, right/arm), a bare name is refused — pass the group path rather than renaming the user's layers.

2. Say what you are about to change

For anything beyond a small touch-up, state the plan in one line and let the user stop you. Especially before: changing the palette, resizing the canvas, merging layers, or editing a layer you did not create.

3. Scope the edit as tightly as you can

  • A selection limits draw, recolor and transform to one area.
  • A layer target keeps you out of everything else.
  • A region on recolor avoids touching the whole cel.

Scoping is cheaper than repairing.

4. Make the change in one call

One draw with all its ops, one recolor. That is one Ctrl+Z for the user if they dislike it. Use label to describe the intent — it becomes their undo entry.

For a pixel-level fix, edit the text, not coordinates: look op="ascii" layer=… rulers=false region=… gives bare rows, the legend and origin; change the wrong cells and send the same rows back as one draw kind grid at that origin. Only that rectangle is touched, and its transparent cells erase — pass transparent:"skip" when you mean to add pixels without clearing any. Glyphs are palette indices, so pieces read separately share one legend: merge their legends when rows from one piece land in another. Only glyphs listed in offPalette (colours outside the palette) can change between reads.

5. Verify what actually changed

look op="diff" fromFrame=… toFrame=…

for animation work, or look op="ascii" on the region for a static edit. A preview shows you the result; a diff shows you the change, including the parts you did not intend. The diff's changedBounds is the quick test: if you edited the hand and the box spans the whole sprite, something else moved.

6. Validate and report

validate

Report what changed and what you deliberately left alone.

Interpreting vague requests

RequestUsually means
"more contrast"Widen the value range of the ramps, not the hue
"cleaner"Remove strays, even out line runs, merge near-duplicate colours
"more detail"Usually wrong at small sizes — ask what should read better
"pop more"Contrast against the background, or a brighter accent
"less flat"Shading exists but does not hue-shift, or has no light direction
"muddy colours"Too many near-duplicates — palette op="extract" on a copy, or analyze + snap
"add a label / score / title"draw kind text; measure with measureOnly first so it lands centred
"closer to the reference / concept"look op="compare", list the three to five largest mismatches, fix only those
"off / uncanny"Run aseprite:review; name the specific cause

Reflect your reading back before acting on it. "More detail" on a 16×16 sprite usually means the sprite needs to be bigger, and that is worth one question.

What not to do

  • Do not redraw a sprite from scratch because it is easier than editing it, unless the user asked for that.
  • Do not "improve" things nobody mentioned.
  • Do not widen the palette to fit a colour you chose. Say the palette lacks it.
  • Do not flatten, resize or re-index without being asked.

Related

aseprite:review, rules://00-core-principles, rules://92-generated-art-tells for what makes art read as machine-made, and the subject file for the part you are fixing (skill://studio, Subject rules) — "fix the hands" starts at rules://34-hands-and-feet.

Frequently asked questions

What does the Fix AI skill do?

Change a sprite that already exists — the user's own work or your earlier output — without destroying what is already right. Use for edits, touch-ups, style corrections and "make it more X" requests.

Why use Fix on TypingMind?

Because you install it once and use it with any model. Fix is plain Markdown rather than provider-specific code, so the same skill runs on GPT-5, Claude, Gemini, Grok, or a local model — and you can switch model mid-chat without it breaking. TypingMind runs on your own API keys, so you pay providers directly instead of a per-seat subscription, and your skills and chats stay in your own storage.

How do I install Fix in TypingMind?

Open Plugins → Skills → Install from GitHub in TypingMind and paste https://github.com/with-pebbly/aseprite-ai-artist/tree/main/skills/fix. TypingMind reads its SKILL.md and installs it as a skill you can enable per chat.

Which AI models can use Fix?

Any model you connect in TypingMind. AI skills are plain Markdown instructions rather than provider-specific code, so GPT, Claude, Gemini, Grok, and local models can all load this skill when a request matches it.

How many AI models can I use with Fix?

As many as you like. As long as a model supports skills, you can use Fix with it — GPT, Claude, Gemini, Grok, DeepSeek, Mistral, Llama and more — all on TypingMind with your own API keys.

Is the Fix AI skill free?

Yes. It is published on GitHub by with-pebbly under the MIT license. You only pay your own AI provider for the tokens you use.

What are AI skills?

An AI skill is a reusable instruction bundle that teaches an AI model how to do one specific task. It follows the open Agent Skills format: a SKILL.md file with a name and description, plus any scripts, templates or reference files the model may need. The model reads the instructions only when your request matches the skill, so an installed skill costs nothing until it is used.

How are AI skills different from plugins or MCP servers?

A plugin or MCP server gives a model new tools to call — code that runs somewhere and returns a result. An AI skill gives the model knowledge and process instead: how to approach a task, which steps to follow, what good output looks like. Skills are plain Markdown, so they need no server, no API key and no runtime, and they work with any model.

View all

Set up your own AI workspace now

Get notified about new features and future giveaways by subscribing to our newsletter 👇