microsoft/cat-agent-skills is a static website that catalogs reusable instruction sets and related packages for AI agents. People use it to search, filter, rate, and download skills for Cowork, Copilot Studio, and Scout, along with Copilot plugins and Scout automations. The catalogue entries are the skills, instructions, plugins, and settings displayed by the site.
Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add microsoft/cat-agent-skills --skill qr-code-buildergit clone --depth 1 https://github.com/microsoft/cat-agent-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/microsoft/cat-agent-skills/qr-code-builder)<a href="https://agentmods.dev/skills/microsoft/cat-agent-skills/qr-code-builder"><img src="https://agentmods.dev/badge/skills/microsoft/cat-agent-skills/qr-code-builder/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/microsoft/cat-agent-skills/qr-code-builder"><img src="https://agentmods.dev/badge/skills/microsoft/cat-agent-skills/qr-code-builder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00176 | $0.05885 |
| Opus 5.5 | $0.00070 | $0.02354 |
| Sonnet 5.5 | $0.00035 | $0.01177 |
| Haiku 4.5 | $0.00018 | $0.00588 |
Grade A, and why
qr-code-builder scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 21d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
| **Safe commands (security)** | Never put a user-derived value — URL, caption, frame text, color, logo path, file name, Wi-Fi / contact / event fields, CSV column names — raw or inside double quotes into a shell string: How it starts
The opening of the file, as written. The whole thing — 239 lines — stays where its author put it; the contents beside it link to each section on GitHub.
QR Code Builder
Make a QR code with as few questions as possible. The only required input is the web address (or the details of a type the user explicitly asked for); every look-and-feel choice is optional and never blocks.
When NOT to use
| Request | Do instead |
|---|---|
| Read, scan or decode an existing QR code | Out of scope (this skill only makes codes) |
| Other barcodes (Code 128, EAN/UPC, Data Matrix, PDF417, Aztec) | Out of scope |
| General illustrations, logos or artwork with no QR code | Use an image-generation tool instead |
| A flyer, poster, page or slide that contains a QR code | Make the QR code here, then use a document or presentation tool for the layout |
| Rebranding documents or decks | Out of scope — "brand colors" here means the QR code's colors only |
| An interactive QR-generator app | Out of scope (this skill delivers files) |
Engine rules
| Rule | Detail |
|---|---|
| Skill folder | <skill-dir> = the folder this SKILL.md was loaded from. Engine: scripts/make_qr.py; style guide: assets/qr-style-guide.png (both relative to <skill-dir>). If unknown, search **/qr-code-builder/scripts/make_qr.py and use its grandparent. Use absolute paths in commands |
| Libraries | Pillow + one QR encoder (ReportLab, qrcode or segno, first found). Never pip install; never write your own encoder. The engine reads CSV only for bulk (XLSX: see Bulk) |
| No separate check | The generation run reports encoder and returns missing_encoder / missing_pillow itself. Run python '<skill-dir>/scripts/make_qr.py' --check ONLY after one of those errors (or another encoder/Pillow error), or on the Customize path to read vector_pdf before answering a PDF question |
| One command | python '<skill-dir>/scripts/make_qr.py' … --scratch '<scratch>' --file-name '<slug>-qr' in ONE call (the engine creates <scratch>; no mkdir). No extra verification calls — the JSON is the check |
| Safe commands (security) | Never put a user-derived value — URL, caption, frame text, color, logo path, file name, Wi-Fi / contact / event fields, CSV column names — raw or inside double quotes into a shell string: $(…), backticks and $VAR still run inside "…". Prefer an argument list when the host runs Python: subprocess.run([sys.executable, script, "--url", url, …]), one list item per value, never shell=True. In a shell, wrap every user value in single quotes and write each ' inside it as '\'' (e.g. Bob's → 'Bob'\''s'). Quote <skill-dir>, <scratch> and file paths the same way. Fixed option names (png, fluid, small, logo) are safe as-is |
| Fresh folder every run | <scratch> = <qr-root>/<slug>-run<N> (Platform notes). Use a NEW folder for EVERY run and rerun — N = 1, 2, 3 …; never reuse one, never write straight into the delivery location. The engine creates it, refuses a non-empty one (scratch_not_empty → next N) and refuses anything outside the working folder (bad_scratch). Never delete files or folders to make room |
| Safe file names | --file-name '<slug>-qr', where <slug> is the web address's host (contoso.com), the code type (wifi, contact, event, text), or for bulk links (bulk file names come from the name column + row number). Never pass a path, folder or a file name the user typed. The engine reduces the name to letters, digits, ., -, _ (it reports the final file_name), and checks every file stays inside <scratch> — quoting alone doesn't stop ../ tricks |
| Images | Standard: never view the PNG. Customize: view the final PNG ONCE, only if the look changed (shape, eyes, frame, logo, colors, caption); after a rerun view only the rerun. The style guide is for the user — don't inspect it |
What ships with it
4 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 21d ago First seen · 239 lines · 176 tokens per session scan A c5e0517d101b
qr-code-builder is a skill published in the GitHub repository microsoft/cat-agent-skills (93 stars, last pushed today), licensed MIT. It adds 176 tokens to every session and 5,885 once invoked, about $0.0007 per session on Opus 5.5. A static security scan graded it A with 1 finding (runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-10-07.
Other skills, from other repositories
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
insight-error-page
Write or audit an insight-kind error page for the Next.js dev overlay. Use when creating a new errors/ .mdx page, auditing an existing one, or checking that a page matches the framework fix cards. Covers page structure, title alignment, FixCard cards with Copy prompt button, code snippets, terminology verification…