Perler Pattern
v1.2.0Convert any image into a perler/hama/fuse bead pattern. 把任意图片转换成精美的拼豆图纸。Use when the user wants to generate a bead pattern from a photo, image, or URL. Outpu...
⭐ 0· 95·0 current·0 all-time
by@tobewin
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
high confidencePurpose & Capability
The name/description (convert images to perler/bead patterns) aligns with the provided instructions and embedded Python script. Required runtime (python3 + common imaging libraries) is appropriate and expected for this functionality.
Instruction Scope
The SKILL.md contains a full Python script that performs offline background removal (GrabCut) and pattern generation, and it instructs the agent to install OpenCV/Pillow/numpy if missing. The instructions do not request unrelated files, secrets, or external endpoints beyond image source URLs supplied by the user. Minor inconsistency: the skill states "fully offline, no API" for processing (true) but the runtime installs use pip which will fetch packages from PyPI (network activity to install dependencies).
Install Mechanism
This is an instruction-only skill with no install spec. It runs in-process pip installs (pip3 install opencv-python-headless Pillow numpy) if missing. Using pip at runtime is standard here but does involve network access to PyPI; no third-party or unusual download URLs are used.
Credentials
The skill declares no required environment variables, no credentials, and no config paths. The embedded code does not reference secrets or unrelated environment variables in the provided excerpt.
Persistence & Privilege
always is false and there is no mechanism to persist or modify other skills or agent-wide settings. The script writes outputs to an OUTPUT_DIR specified by the agent/user (normal for a generator).
Assessment
This skill appears coherent and implements what it claims, but take the usual precautions before running code that will install packages and process files: 1) Review the provided Python script (it will run on your system). 2) Run it in a controlled environment or virtualenv to avoid unexpected system-wide package changes. 3) When supplying OUTPUT_DIR, choose a dedicated directory (avoid system paths). 4) If you provide an image URL, ensure it’s from a trusted source (the skill may fetch that URL). 5) Expect pip to download OpenCV/Pillow/numpy from PyPI at runtime — this is normal but does require network access. If any of these behaviors are unacceptable, don’t run the script or request the maintainer to provide a packaged install or explicit dependency list.Like a lobster shell, security has layers — review code before you run it.
latestvk97e8sr4zrc6qyfdmph5qek89h83acz4
License
MIT-0
Free to use, modify, and redistribute. No attribution required.