| name | slim-badges |
| description | Add badges to a repository, README, or document. Ships with authorship badges (100% Human Made, AI-assisted, AI-generated, AI-autonomous) and the SLIM Best Practices badge, and is extensible to any badge. Use when a user wants to add, choose, or generate a badge for a project or document. |
Badges
Overview
A skill for adding badges to a repository, README, or any document. Badges are
small status/label images (rendered via shields.io) that
communicate something at a glance — how a document was authored, what standards a
project follows, and more.
The skill ships with a small, extensible catalog (see assets/badges.json).
The starter badges are:
Authorship — how a document was written, along the human↔AI spectrum:
| Badge | Meaning | When to use |
|---|
 | Written entirely by a human, no AI generation | Hand-written docs, essays, posts you authored yourself |
 | Human-authored with AI help (drafting, editing, suggestions) | You wrote it but used an AI assistant along the way |
 | Primarily produced by an AI model | Content generated by an LLM with little human authorship |
 | Produced by an AI system acting autonomously | Content from an autonomous/agentic AI with no human in the loop |
SLIM — badges for crediting the SLIM project:
| Badge | Meaning | When to use |
|---|
 | Project follows SLIM best practices | Projects adopting SLIM guidance |
When to Use This Skill
Trigger when the user mentions adding or choosing a badge / shield for a
repo or document, wants to disclose authorship (human vs. AI), or asks for the
SLIM badge. Keyword cues: badge, shield, slim badge, human made,
AI-assisted, AI-generated, AI-autonomous, best practices.
Workflow
Step 1: Identify which badge(s) the user wants
Read the user's request and scan for keywords, matching against the catalog in
assets/badges.json (each entry has a keywords list):
- "human made / human written / no AI" → 100% Human Made
- "AI-assisted / wrote it with AI / copilot" → AI-assisted
- "AI-generated / generated by AI / LLM-written" → AI-generated
- "AI-autonomous / agentic / no human in the loop" → AI-autonomous
- "slim / best practices" → Best Practices from SLIM
If exactly one badge clearly matches, proceed. If the request is ambiguous or
matches none/several, ask the user which badge they want, presenting the catalog
options.
Step 2: Identify the target document
Determine where the badge goes (e.g. README.md, a specific doc, a badges row near
the title). If unclear, ask. Check for an existing badges row to append to rather
than duplicating.
Step 3: Insert the badge markdown
Each badge is a linked image. Always use the absolute shields.io URL (the
catalog's image_url) so the badge renders everywhere — including where the
SLIM marketplace renders this SKILL.md (relative paths resolve against the page URL
and break). Use the catalog's alt text so the badge is accessible. Ready-to-paste
snippets for all badges:
[](https://nasa-ammos.github.io/slim/?search=Badges)
[](https://nasa-ammos.github.io/slim/?search=Badges)
[](https://nasa-ammos.github.io/slim/?search=Badges)
[](https://nasa-ammos.github.io/slim/?search=Badges)
[](https://nasa-ammos.github.io/slim/)
Place badges at the top of the document (just under the title) where badges
conventionally live.
Offline / self-hosted option: static SVG snapshots of each badge are bundled in
assets/ (e.g. assets/human-made.svg). Copy one into the target repo and
reference it locally if you prefer not to depend on shields.io at render time.
Badge Catalog Details
| id | Label | Color | Link |
|---|
human-made | 100% Human Made | green | skill page* |
ai-assisted | AI-assisted | gold | skill page* |
ai-generated | AI-generated | red | skill page* |
ai-autonomous | AI-autonomous | red | skill page* |
slim-best-practices | Best Practices from SLIM | blue | https://nasa-ammos.github.io/slim/ |
* skill page = https://nasa-ammos.github.io/slim/?search=Badges
The authorship badges use a traffic-light color scheme (green → gold → red) to
read at a glance from fully human to fully AI; AI-generated and AI-autonomous share
red as the most-AI end of the spectrum.
Naming & Wording
Badge labels follow title case, but keep the element after a hyphen lowercase
unless it is an acronym or proper noun — e.g. Human-made, AI-assisted (AI is
an acronym → caps; assisted → lowercase), AI-generated, AI-autonomous. Keep
labels short; put the fuller explanation in the badge's alt text.
Extending the Catalog
To add a new badge:
- Append an entry to
assets/badges.json with: id, category, label, alt,
color, image_url, link_url, snapshot, keywords, and description.
- Compose the shields.io URL with the query form (hyphens are literal here — no
escaping needed):
https://img.shields.io/static/v1?label=&message=<Label>&color=<color>
Pick a named color (green, gold, red, blue, …) or a hex code.
- Reference the badge with the absolute
image_url and the alt text in
markdown so it renders everywhere and stays accessible (relative paths break when
the marketplace renders SKILL.md).
- Optionally cache an SVG snapshot:
curl -fsSL "<image_url>" -o assets/<snapshot>.
Assets Available
badges.json — the badge catalog (source of truth).
human-made.svg, ai-assisted.svg, ai-generated.svg, ai-autonomous.svg,
slim-best-practices.svg — static SVG snapshots of each badge.
Credits
The human/AI authorship-badge concept was inspired by
byhumannotai.com.