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 own-voice-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/own-voice-builder)<a href="https://agentmods.dev/skills/microsoft/cat-agent-skills/own-voice-builder"><img src="https://agentmods.dev/badge/skills/microsoft/cat-agent-skills/own-voice-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/own-voice-builder"><img src="https://agentmods.dev/badge/skills/microsoft/cat-agent-skills/own-voice-builder.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector warn
SkillSpector: 1 finding, up to high
These are SkillSpector’s own severities. On a checked sample its high-severity flags on skills were ~96% false positives — a documented command, a public API, a “never do X” rule — so we show them as a caution to read, not a verdict. Why →
- high System Prompt Leakage · line 55 Skill contains instructions that could directly expose system prompts, internal rules, or hidden instructions to users or external parties.Fix: Remove any instructions that reveal, print, or output system prompts or internal rules. System instructions should never be exposed to end users.
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 | $0.00104 | $0.01628 |
| Opus 5 | $0.00052 | $0.00814 |
| Sonnet 5 | $0.00021 | $0.00326 |
| Haiku 4.5 | $0.00010 | $0.00163 |
Grade A, and why
own-voice-builder scanned grade A with 0 findings 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 6d 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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 71 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Own Voice Builder
Follow the phases below to analyze the user's sent emails and generate (or update) a personal voice skill (default name: <firstname>-voice).
Phase 0 — Setup questions (keep it short)
Ask the user, in one message, only what you cannot infer:
- Time window: How far back to analyze (default: last 12 months).
- Exclusions: Any recipients, domains, or topics to skip (e.g. HR, legal, personal contacts).
- Skill name: Default
<firstname>-voice; confirm the first name.
If the user already stated any of these in the conversation, do not re-ask. Check the user's existing memories/preferences for style rules (e.g. "no em-dashes", "never start with 'I hope this finds you well'") — these become hard rules in the generated skill, no exceptions.
Phase 1 — Sample & analyze sent mail
Sampling (do this in parallel where the harness allows):
- Use the mail tool available in this harness (e.g.
m365_list_emailswithfolder="sent") to pull 100–150 sent emails, stratified across the whole time window in 4–6 time buckets so the sample isn't biased toward recent weeks. - Retrieve
subject,toRecipients,sentDateTime,bodyPreview, and fullbodywhere needed. - Classify each message: internal vs. external recipient (by domain), language, one-to-one vs. group, thread reply vs. fresh mail.
- Skip anything matching the user's exclusions. Skip auto-generated mail (calendar responses, forwards without added text, one-word replies) — they carry no style signal.
Analyze the sample for:
- Language routing: which languages the user writes in (one, several, or mixed within a mail) and what triggers each — recipient domain, thread language, specific people. Make no assumptions about which languages these are; detect them from the sample.
- Register modes: distinct tone modes (e.g. formal-external, warm-internal, terse-lowercase-quick-reply) and the contexts that trigger each.
- Structural fingerprints: greetings, openers, how they get to the point, closing patterns, sign-offs — per mode and per language.
- Vocabulary: recurring connectors, action verbs, hedges, signature phrases.
- Punctuation & formatting habits: em-dash usage, emoji frequency and which ones, exclamation marks, date formats, bullet style, paragraph length.
- Behavioral patterns: how they decline, propose meeting times, escalate, follow up, deliver bad news.
- Evolution: if the style changed over the window, weight recent months higher.
- Taboo list: words, phrases, and patterns that never appear in their mail — including typical AI-isms absent from it (English examples: "I hope this email finds you well", "delve", "leverage"; identify the equivalent stock phrases in each of the user's languages).
What ships with it
1 file 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.
- 6d ago First seen · 71 lines · 104 tokens per session scan A 2931ba3b66ec
own-voice-builder is a skill published in the GitHub repository microsoft/cat-agent-skills (66 stars, last pushed 2d ago), licensed MIT. It adds 104 tokens to every session and 1,628 once invoked, about $0.0005 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-03.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
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…
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…
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…
next-cache-components-optimizer
Drive a Next.js route to instant navigation by setting up an agentic loop, under Cache Components / PPR, on initial load (hard navigation) and client-side navigation (soft navigation). Encode the goal as a failing @next/playwright instant() e2e and work it to green, one verified route at a time; the shipped test then…
next-partial-prefetching-adoption
Turn on Partial Prefetching in a Next.js app and work through the insights it surfaces. Use when the user wants to enable or adopt Partial Prefetching, flip the partialPrefetching flag, opt routes in with export const prefetch = 'partial', audit Link prefetch={true} behavior, preserve existing prefetched UI with…