Skip to main content
Glama
aseprite-claude-connector

aseprite-claude-connector

pixelate_reference

Converts a reference image into pixel art on a new layer by averaging pixels, reducing to a chosen palette, with optional dithering, despeckling, background removal, and edge cleanup.

Instructions

Turn a reference picture into pixel art on a new layer: the faithful way to recreate a photo or painting. The picture (or the crop_* part of it, in its own pixels) is shrunk to width x height by averaging, so each pixel is the true colour of that area. height defaults to the picture's aspect. It is then snapped to colors colours chosen from the picture, or to palette (#rrggbb list) to match an existing one. dither is none, ordered, or diffusion. despeckle removes lone noise pixels. remove_background clears a white sheet or flat backdrop (and the light fringe it leaves), before the palette is chosen so no colours are spent on it. clear_edges drops bits of neighbouring art a crop catches at its edges, keeping the subject and anything floating clear of the edges. To cut every pose out of a character sheet at once, use split_sheet. Refine the result by hand, and check it with compare_to_reference. One undo step. Headless: pass file.

Tool call template

Replace each placeholder with a value matching the input schema. Only required parameters are included; add optional parameters as needed.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "image_path": "<image_path>",
      "width": "<width>"
    },
    "name": "pixelate_reference"
  }
}

Input Schema

NameRequiredDescriptionDefault
xNo
yNo
fileNo
widthYes
colorsNo
crop_xNo
crop_yNo
ditherNonone
heightNo
paletteNo
despeckleNo
crop_widthNo
image_pathYes
layer_nameNoReference pixels
clear_edgesNo
crop_heightNo
remove_backgroundNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.4.0

TDQS

A4.3/5.0
Behavior4/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations contain only a title, so the description carries the full disclosure burden and does substantial work: it states output goes to a new layer, that the result costs "one undo step" (reversibility), that width/height downsampling is by area averaging, that remove_background runs before palette selection so no colours are spent on the backdrop, and that clear_edges preserves edge-floating subject pixels. It omits any statement about required permissions, failure modes on missing files, and whether the new layer becomes active.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

The core action, sizing, and colour reduction are front-loaded in the first three sentences, and later sentences are additive rather than repetitive. It is dense but each clause carries distinct information; the backtick-quoted parameter names make scanning easy despite the single-block paragraph format.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness4/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a 17-parameter destructive-ish image operation with no annotations and no output schema, the description covers process, ordering effects, undo cost, and the headless path. What is missing is the meaning of x/y and layer_name and any hint of what the call returns, which matters since no output schema exists to explain it.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

With 0% schema description coverage, the description must compensate and mostly does: it defines width, height (defaults to the picture's aspect), colors (palette entry count), palette as a #rrggbb list, and crucially enumerates dither's three legal values (none/ordered/diffusion) even though the schema declares no enum, and explains despeckle, remove_background, clear_edges, the crop_* family (in the picture's own pixels), and file for headless use. Three parameters remain unexplained — x, y, and layer_name — leaving room above a 4.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The opening sentence gives a specific verb and resource ("Turn a reference picture into pixel art on a new layer") and immediately frames the tool's niche as "the faithful way to recreate a photo or painting." That framing distinguishes it from compare_to_reference (verification) and split_sheet (sheet cutting), both of which are named later, so an agent can place it without opening any schema.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines4/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It routes explicitly to alternatives with conditions: split_sheet when the goal is cutting every pose out of a character sheet, compare_to_reference when the goal is verifying the output. It also signals an implicit workflow (pixelate, then refine by hand, then compare). It stops short of stating when this tool should not be used, e.g. for existing pixel art that needs no downsampling.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Deploy Server

Other Tools