A collection of badge assets for AWS and Anthropic certifications — vector templates for editing and raster examples for reference.
A small Python starter for AWS and Anthropic badges. It automates preparation; tracing and final typography stay in Inkscape. Existing assets remain untouched.
Install mise, then run:
mise trust
mise install
just setup
uv run badge-prep --helpIf mise is not activated in your shell, run mise exec -- just setup and prefix
subsequent commands with mise exec -- too.
Prepare a single image or every supported image directly inside a directory:
uv run badge-prep prepare "assets/raster examples" --output output/aws --size 1024 --padding 32 --colors 8source accepts one supported image or a flat directory. Supported formats are PNG,
JPEG, WebP, TIFF, and BMP. Defaults are --output output, --size 1024,
--padding 32, and --colors 8; size is limited to 1–8192 pixels, padding must
leave room on the canvas, and colors is limited to 1–256.
Each input produces <stem>.png and <stem>.palette.json. Preparation crops only
transparent margins, preserves aspect ratio, and centers the result on a transparent
square. Opaque backgrounds stay intact. Palette JSON lists approximate RGB colors
from the original visible pixels, most frequent first; partially transparent pixels
count once using their underlying RGB values. Gradients and antialiasing affect matches.
The PNG retains the original colors rather than applying the palette.
The palette JSON schema has a source filename string and a colors array of
six-digit hexadecimal RGB strings. It describes approximated colors found in the
source image; it does not include transparent pixels or change the prepared PNG.
Only the first frame is used, after EXIF orientation is applied. Directory traversal is nonrecursive. Existing outputs, duplicate stems, symlink inputs, and output directories nested inside the batch input are rejected before writing. Each PNG and palette pair is staged in the destination directory and published with hard links; reported publish errors trigger rollback attempts. The destination filesystem must support hard links. Two output files cannot be made visible as one transaction, so abrupt process termination between publishes can leave one file. Batch members are processed independently: failures are reported with their source paths, successful pairs are kept, and the command exits 1 if any member fails. Invalid command arguments exit 2. Choose a new output directory for another run.
Choose a text rectangle in the original image using an image editor, then export it:
uv run badge-prep font-sample "assets/raster examples/aws-certified-cloud-practitioner-logo-600px.png" --box 100,250,500,350 --output output/text-sample.png --scale 3 --background whiteThose coordinates illustrate the syntax; choose the bounds around your target text.
Coordinates are left,top,right,bottom in pixels after EXIF orientation, with right
and bottom excluded. Export one text style at a time and retain distinctive letters.
The source must be one supported raster image. Defaults are output
output/font-sample.png, --scale 3, and --background white; scale is limited to
1–8. The output path must not already exist. Choose a background that contrasts with
the text; the option fills transparent pixels and does not remove an existing background.
Enlargement adds no missing image detail. Invalid crop bounds, input data, or write failures
exit 1; invalid command arguments exit 2.
The helper saves an RGB PNG and prints links to WhatTheFont and Font Squirrel Matcherator. Upload the crop manually, compare candidate letter shapes, then confirm the font visually in your SVG. These services provide candidates, not guaranteed matches; custom logo lettering may have no equivalent font. Font selection, acquisition, and licensing remain manual. The CLI makes no network calls or automatic uploads.
just ci
just fmt
just lint
just typecheck
just test
just buildjust ci runs the same lint, build, type-check, and test stages as GitHub Actions.
Python 3.14 runs the CLI and tests. just setup installs the pinned tools, syncs
locked dependencies, and installs pre-commit and commit-message hooks.
uv locks dependencies; hatchling builds the package. Typer/Rich provide the CLI,
Pillow handles images, Ruff runs through prek, Pyright uses strict mode, and pytest
checks image behavior and output protection. External tool versions and the
virtual environment live in mise.toml; Python tools are declared in pyproject.toml and resolved in uv.lock. Run uv lock
after changing dependencies and include the updated lockfile in the same change.
Run just to list commands. Generated files belong under ignored output/.
Project authoring rules live in docs/ai-instructions.md.
All linters and formatters run through prek. Hooks cover Python linting and
formatting, strict types, Markdown, TOML, YAML, spelling, file integrity, and
Conventional Commit messages. just lint skips Pyright because just typecheck
owns that check; commits still run its hook. just fmt formats Python and TOML.
just lint also runs actionlint on GitHub workflows. Source assets remain
untouched; the large-file guard allows files up to 5 MiB.
The supplied Anthropic raster acceptance results, source hashes, and manual inspection notes are recorded in the badge preparation acceptance record.
Spelling rules and exclusions live in .cspell.config.yaml; project words live in
.cspell/project.txt. Shared VS Code settings select the project interpreter,
pytest, Ruff, and that dictionary; extension recommendations are included.
| Preview | Certification | Format | File |
|---|---|---|---|
 |
AWS Certified (Generic Template) | SVG | aws-certified-template.svg |
 |
Anthropic Certified (Generic Template) | SVG | anthropic-certified-template.svg |
 |
AWS Certified AI Practitioner | SVG | aws-certified-ai-practitioner.svg |
 |
AWS Certified Cloud Practitioner | SVG | aws-certified-cloud-practitioner.svg |
 |
AWS Certified Solutions Architect Associate | SVG | aws-certified-solutions-architect-associate.svg |
 |
AWS Certified Solutions Architect Professional | SVG | aws-certified-solutions-architect-professional.svg |
 |
Claude Certified Architect Professional | SVG | claude-certified-architect-professional.svg |
 |
Claude Certified Architect Foundations | SVG | claude-certified-architect-foundations.svg |
 |
Claude Certified Associate Foundations | SVG | claude-certified-associate-foundations.svg |
 |
Claude Certified Developer Foundations | SVG | claude-certified-developer-foundations.svg |
assets/
├── vector templates/
│ ├── anthropic-certified-template.svg # Editable SVG base template
│ └── aws-certified-template.svg # Editable SVG base template
├── vector rendered/
│ ├── aws-certified-ai-practitioner.svg
│ ├── aws-certified-cloud-practitioner.svg
│ ├── aws-certified-solutions-architect-associate.svg
│ ├── aws-certified-solutions-architect-professional.svg
│ ├── claude-certified-architect-professional.svg
│ ├── claude-certified-architect-foundations.svg
│ ├── claude-certified-associate-foundations.svg
│ └── claude-certified-developer-foundations.svg
└── raster examples/
├── aws-certified-ai-practitioner-logo-600px.png
├── aws-certified-cloud-practitioner-logo-600px.png
├── aws-certified-solutions-architect-associate-logo-600px.png
└── aws-certified-solutions-architect-professional-logo-600px.png