io.github.dengyu123456/crochet-engine

Official registry Media and design

Deterministic photo-to-crochet-chart engine: stitch counts, 5 techniques, yarn estimates.

Install

Add this to your client's MCP configuration (Claude Desktop: claude_desktop_config.json; Cursor: .cursor/mcp.json; VS Code and others use the same "mcpServers" shape).

PYPI package crochetpatterngen-engine
Claude Code
claude mcp add crochet-engine -- uvx crochetpatterngen-engine==0.1.1 --from 'crochetpatterngen-engine[mcp]' crochetpatterngen-mcp
JSON configuration (Claude Desktop, Cursor, VS Code)
{
    "mcpServers": {
        "crochet-engine": {
            "command": "uvx",
            "args": [
                "crochetpatterngen-engine==0.1.1",
                "--from",
                "crochetpatterngen-engine[mcp]",
                "crochetpatterngen-mcp"
            ]
        }
    }
}

Documentation

From the project's README on GitHub, refreshed weekly. Links open on GitHub.

Contents
  1. crochetpatterngen-engine
  2. Why this engine
  3. Install
  4. Python API
  5. CLI
  6. MCP server (Claude Desktop & friends)
  7. ASCII design format
  8. Development
  9. License

crochetpatterngen-engine

The deterministic photo-to-crochet-chart engine behind crochetpatterngen.com — packaged as a Python library, CLI, and MCP server so AI agents (Claude Desktop, Cursor, etc.) can turn photos into crochet charts with exact stitch counts.

Deterministic, not generative. There is no AI image generation here — just resize, median-cut color quantization, and confetti cleanup. Same input, same chart, same stitch counts, every time. No hallucinated rows.

Why this engine

  • Real stitch counts — every chart comes with per-color hex + stitch counts that sum exactly to the grid size. What you see is what you crochet.
  • C2C block aspect correction — C2C blocks are physically wider than tall (0.7:1 h:w). The engine adjusts grid dimensions so the worked piece keeps the photo's proportions instead of coming out squashed.
  • 5 technique presets — graph, c2c, tapestry, mosaic, filet.
  • Yarn estimation — per-color yardage for 6 yarn weights (fingering → super bulky).
  • Zero heavy deps — Pillow + numpy only. Optional rembg extra for background removal.

Install

pip install crochetpatterngen-engine          # core
pip install crochetpatterngen-engine[mcp]     # + MCP server
pip install crochetpatterngen-engine[rembg]   # + background removal

Python API

from PIL import Image
from crochet_chart_engine import generate_chart, render_design, estimate_yarn

png_bytes, meta = generate_chart(
    Image.open("photo.jpg"),
    technique="c2c",     # graph | c2c | tapestry | mosaic | filet
    grid_w=60,           # 20..120
    n_colors=8,          # 2..16
)
# meta: {"grid_w", "grid_h", "colors": [{"hex", "count"}],
#        "total_stitches", "technique", "warning"}

yarn = estimate_yarn(meta, weight="worsted")
# {"total_yards": ..., "colors": [{"hex", "count", "yards"}], ...}

png_bytes, meta = render_design("designs/frog.txt")  # ASCII design -> chart

CLI

crochet-chart chart photo.jpg out.png --technique c2c --grid-w 60 --colors 8
crochet-chart design designs/frog.txt frog.png
crochet-chart yarn photo.jpg --technique graph --weight worsted

MCP server (Claude Desktop & friends)

Add to claude_desktop_config.json:

{
  "mcpServers": {
    "crochet": {
      "command": "uvx",
      "args": ["--from", "crochetpatterngen-engine[mcp]", "crochetpatterngen-mcp"]
    }
  }
}

Or with a regular pip install:

{
  "mcpServers": {
    "crochet": {
      "command": "python",
      "args": ["-m", "crochet_chart_engine.mcp_server"]
    }
  }
}

Tools exposed (stdio transport):

ToolWhat it does
photo_to_chartLocal image path → chart PNG (base64 or saved to output_path) + per-color stitch counts JSON
render_ascii_designASCII design file → chart PNG + stitch counts
estimate_yarn_toolStitch counts → per-color yardage for one yarn weight or all

ASCII design format

See designs/frog.txt and designs/dragon.txt:

# comment
palette: G=#58A05C, W=#FFFFFF, .=#FAFAF7
title: Frog Face
technique: graph
---
..GGGG..........GGGG....
.GGWWGG........GGWWGG...

One character per stitch; . is the background/padding color. Rows are padded to the longest row.

Development

pip install -e .[dev,mcp]
pytest tests/

License

MIT — see LICENSE.


Free web tool + pattern library at crochetpatterngen.com.

Frequently asked questions

What is io.github.dengyu123456/crochet-engine?

Deterministic photo-to-crochet-chart engine: stitch counts, 5 techniques, yarn estimates.

How do I install io.github.dengyu123456/crochet-engine?

It runs on your machine. In Claude Code: claude mcp add crochet-engine -- uvx crochetpatterngen-engine==0.1.1 --from 'crochetpatterngen-engine[mcp]' crochetpatterngen-mcp. For other clients, paste the JSON configuration from this page into their MCP settings.

Does io.github.dengyu123456/crochet-engine need an API key?

Its registry entry lists no required key or secret. Some features may still need an account with the service it connects to.

Is io.github.dengyu123456/crochet-engine open source?

Yes. Its code is on GitHub under the MIT license.

Who maintains io.github.dengyu123456/crochet-engine?

It is published by dengyu123456. AI.tn lists it but does not maintain or endorse it.

From the official MCP Registry. Listing is not an endorsement: review a server's code and permissions before giving it access to your data.