Skip to main content
Glama
aseprite-claude-connector

aseprite-claude-connector

compare_to_reference

Read-only

Compares the canvas to a reference image and returns match scores, side-by-side views, and a heatmap highlighting the worst areas to fix first.

Instructions

See how close the canvas is to a reference picture. Returns the reference shrunk to the canvas size, the canvas, and a heatmap (red = most different) side by side, plus scores out of 100: score per pixel, shape_score with fine detail blurred away (composition, values, and colour masses), and close_pixels (% that nearly match). worst_areas are the cells of a grid x grid split that differ most, with both average colours and whether the art is darker or lighter: fix those first. crop_* picks the part of the reference the canvas shows. layer compares one layer; clear pixels count as black. 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>"
    },
    "name": "compare_to_reference"
  }
}

Input Schema

NameRequiredDescriptionDefault
fileNo
gridNo
frameNo
layerNo
crop_xNo
crop_yNo
max_sizeNo
crop_widthNo
image_pathYes
crop_heightNo

Schema Changelog

Changes observed during successful MCP inspections.

  1. First observedv0.4.0

TDQS

A3.9/5.0
Behavior4/5

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

Annotations only declare readOnlyHint=true, and the description genuinely adds to that: it enumerates the returned artifacts, defines each score's meaning (per-pixel vs blurred-detail vs near-match percentage), explains that worst_areas includes average colours and a darker/lighter flag, and notes the headless path via 'file'. It does not discuss cost, resolution limits, or failure modes for a missing/unreadable reference, so it stops just short of a 5.

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 output is front-loaded with the core action and then the return payload, so an agent gets the essentials in the first sentence. It is dense and somewhat telegraphic (backtick-heavy, long run-on middle), but nearly every clause carries information the caller needs given the absent output schema.

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 10-parameter, read-only analysis tool with no output schema, the description does the important work of documenting the return shape and the scoring semantics. It falls short only on the file-vs-image_path invocation distinction, frame, and max_size, which an agent would have to guess at.

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

Parameters3/5

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

With 0% schema description coverage across 10 parameters the description has to carry the load, and it partially does: grid is defined as the 'grid x grid' split, crop_* is described as selecting the region of the reference, layer explains that clear pixels count as black, and file is tied to headless use. However, the relationship between the required image_path and file is never stated, and frame and max_size are completely unaddressed.

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 first sentence gives a specific verb and resource ('See how close the canvas is to a reference picture') and is followed immediately by exactly what is produced (side-by-side reference/canvas/heatmap plus three scores). An agent can distinguish this from pixelate_reference or add_reference_image purely from the text, since this is the only one framed as a measurement.

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

Usage Guidelines3/5

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

Usage is implied rather than stated: the tool clearly belongs after a reference image exists and before corrective edits ('fix those first'), but there is no explicit when-to-use/when-not-to-use guidance and no named alternative to call instead. The workflow hint is helpful but stops short of real routing guidance.

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