skills.rest
caidish
caidishCOMMUNITY
Explore all skills by this author →
GitHub

nanodevice_flakedetect

Orchestrates sub-skills to detect van der Waals flakes and commit polygons to KLayout.

32|6|Updated Mar 8, 2026
Search Tags:#orchestration#detection#flakes#image-processing#klayout#vdw-heterostructures
One-click install
npx skills add https://github.com/caidish/KlayoutClaw --skill nanodevice-flakedetect
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: nanodevice_flakedetect
Source: https://github.com/caidish/KlayoutClaw/tree/main/skills/nanodevice_flakedetect
Command: npx skills add https://github.com/caidish/KlayoutClaw --skill nanodevice-flakedetect

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill requires opencv-python, numpy, scipy, and includes scripts (resource) components.

What problem does it solve?

Orchestrates a five-subskill pipeline to detect and map van der Waals heterostructure flakes from microscope images and commit them as polygons to KLayout, reducing manual supervision.

Core Features & Use Cases

  • End-to-end stack detection: coordinates align, detect, combine, commit, and review steps automatically.
  • Subskill orchestration: dispatches dedicated SKILL.md based sub-skills for each stage (align, detect, combine, commit, review).
  • Input gathering & validation: collects source images, pixel size, and output directory; handles mirroring for PDMS transfers.
  • Output artifacts: generates detection masks, transformed traces, overlay images, and final polygons in KLayout.

Quick Start

Dispatch nanodevice_flakedetect to run the end-to-end stack detection workflow on your source images and commit the resulting polygons to KLayout.

Frequently Asked Questions about nanodevice_flakedetect

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I automate van der Waals flake detection from microscope images?▼

Automate van der Waals flake detection by orchestrating a pipeline that aligns source images, detects flakes, combines masks, and commits polygons to KLayout. This workflow processes multiple microscope images to generate per-material masks and overlay diagnostics automatically.

How do I map detected 2D material flakes as polygons in KLayout?▼

Map detected 2D material flakes in KLayout by running an automated detection pipeline that transforms image traces into polygons. The workflow commits these polygons directly to KLayout, generating overlay diagnostics for verification.

Does this flake detection workflow handle image alignment for PDMS transfers?▼

The flake detection workflow handles image alignment for PDMS transfers by accepting an optional mirror flag during input gathering. It aligns multiple source images precisely before detecting flakes and generating per-material masks.

What inputs do I need to run end-to-end vdW heterostructure flake detection?▼

End-to-end vdW heterostructure flake detection requires source image paths and pixel size as inputs, along with an optional mirror flag. The workflow uses these inputs to execute alignment, detection, and polygonization steps with deterministic outputs.

Can OpenCV and Python be used for van der Waals flake image processing?▼

OpenCV and Python are used for van der Waals flake image processing within this detection workflow. The pipeline leverages opencv-python, numpy, and scipy to process microscope images, generate detection masks, and produce transformed traces.

Related Skills

Explore More Software Engineering Skills→
opmauopmau
6

check-sizes

Check source file sizes against CLAUDE.md limits and report violations.

Community
Intermediate
caidishcaidish
32

nanodevice_flakedetect_review

Coordinate visual QA for nanodevice flake polygons in KLayout.

Community
Intermediate
caidishcaidish
32

nanodevice_flakedetect_commit

Insert detected flake polygons from traces.json into KLayout as GDS geometry.

Community
Intermediate
caidishcaidish
32

klayoutclaw:geometry

Draw and place rectangles, polygons, paths, cells, and cell instances in KLayout via MCP scripts.

Community
Intermediate