GetVect skill
GetVect turns PNG, JPEG and BMP into SVG, EPS, DXF and PDF locally; its skill teaches an agent to call the command-line tracer.
- 103 ★ on the repository
- Last commit
- MIT
Minimal access
Why this level
- The skill runs a local program and writes files; AI Enhance sends the image to Google but is off by default.
Install
- Works in
In your terminal, with SkillFoxx CLI
npx skillfoxx add skills/getvectDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewMinimal accessThe tool runs nothing and makes no network calls.
- The skill runs a local program and writes files; AI Enhance sends the image to Google but is off by default.
This is third-party code. Review the repository files before installing.
What it does
GetVect is an Electron desktop app and a browser version that convert raster images into vector paths with no server upload, account or per-file fee. Settings cover palette size, detail, smoothing, speck removal, the black and white threshold and roundness. For agents the repository holds `SKILL.md`: the skill describes the command-line call (`node bin/getvect.mjs`) and says when tracing fits. Good targets are logos that must scale, line art for cutters and CAD (`dxf`) and flat illustrations with editable color layers. It excludes photographs and cases that only need a different size. Tracing makes no network calls, and the optional AI Enhance runs only on desktop with your own key.
Who it is for. Designers and engineers whose agents generate raster logos and diagrams that must be made scalable or sent to a cutter.
Good fit when
- A logo generated by an agent must be obtained as SVG.
- Line art is prepared for a plotter or machine in DXF.
- A PNG diagram must be embedded in a document at any size.
- The image holds confidential material and must not go to online vectorizers.
Not a fit when
- A photographic rendition of a picture must be preserved.
- Simply resizing the image is enough.
- Exact lossless reproduction of fine detail is required.
Example request
Convert logo.png to SVG with a six-color palette and put a logo.dxf for the machine next to it.Limitations
The agent path targets macOS and Windows; console use without the app needs a Node build (`npm run build:node`). Through Homebrew the app builds from source on your machine, which takes minutes. AI Enhance sends the image to Google under your key, desktop only. A once-per-launch update check contacts GitHub and is disabled with `GETVECT_NO_UPDATE_CHECK=1`. Tracing does not preserve photorealism because it turns colors into a limited palette.
How to disable. Delete the `getvect` skill from the agent's skills directory, and the app: drag `GetVect.app` out of Applications or run `brew uninstall getvect`. Settings are removed by hand.
FAQ
How do I install on a Mac?
Run `brew install craigjmidwinter/tap/getvect`. It builds from source and takes a few minutes.
Which formats go in and out?
PNG, JPEG and BMP in. SVG, EPS, DXF, PDF and PNG out.
Do images go over the network?
Not when tracing. There are two network touchpoints: optional AI Enhance with your key and an update check on GitHub.
Where is the command-line reference?
In `docs/CLI.md`: flags with ranges, exit codes and failures; `SKILL.md` is written for the agent.
Is there a version without installing?
Yes, in the browser at getvect.midwinter.io/app: after loading, the page runs entirely on your machine.
Discussion
No questions or reviews yet.
Related
SkillsA skill that draws interactive isometric illustrations in SVG, with WebGL shaders and React output, explaining an idea as a working machine
An MCP server that drives Inkscape through conversation: it edits an open document live on the canvas or processes SVG headlessly.
A skill hand-writes SVG feature illustrations by one SPEC.md ruleset, with a single palette serving both light and dark themes.
Skills for an agent that makes video, images, speech and music through the PixVerse CLI using ready pipelines.
Vane GPT Image Skills: two Codex skills for GIF sticker packs from a photo and pixel covers with illustrations for X articles
Pixelbolt lets an agent draw and edit pixel art from a JSON recipe: sprites, effects and animations as PNG, GIF and atlases.
README in short
A local vectorizer as an app and a command line with an agent skill. It works offline with no per-file fee.