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 draft-new-contractgit 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/draft-new-contract)<a href="https://agentmods.dev/skills/microsoft/cat-agent-skills/draft-new-contract"><img src="https://agentmods.dev/badge/skills/microsoft/cat-agent-skills/draft-new-contract/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/draft-new-contract"><img src="https://agentmods.dev/badge/skills/microsoft/cat-agent-skills/draft-new-contract.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 | $0.00151 | $0.03172 |
| Opus 5.5 | $0.00060 | $0.01269 |
| Sonnet 5.5 | $0.00030 | $0.00634 |
| Haiku 4.5 | $0.00015 | $0.00317 |
Grade A, and why
draft-new-contract 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 2d 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 — 237 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Draft new contract
Turn one precedent agreement into a new working draft, and optionally fill that draft in. These are two separate phases, and conflating them is the main way this goes wrong.
Phase 1 — Prepare. Convert the precedent into a new .docx in which every
party-specific and transaction-specific detail is replaced by a visible
placeholder. This needs only the precedent — not the deal facts, the
counterparty, or any values. It is a complete deliverable on its own:
"templatise this" or "strip the counterparty out" wants the placeholdered
document and nothing more. Deliver it and stop.
Phase 2 — Populate. Fill placeholders with grounded values. Optional, repeatable, frequently a later conversation. The user filling the document in themselves is a normal outcome, not a failure.
The phases may run back-to-back in one turn ("draft an NDA for Acme Trading based on this one") or weeks apart. Never block phase 1 waiting for information that belongs to phase 2. If you have a precedent, you can produce the draft.
Three rules frame everything below. The deliverable is always a Word document — a contract that exists only as chat text cannot be circulated, redlined or signed. The precedent is never modified — you always produce a new document. You draft, you do not advise — where the precedent's position is a choice rather than boilerplate, carry it across and flag it for decision; never substitute your own judgement silently.
When NOT to Use
- Deriving standard positions or negotiation guidance from a body of approved
precedent — use
create-playbook. That skill synthesises review guidance from MANY precedents; this one produces ONE new agreement from ONE precedent. - Questions about what a precedent says, or narrative comparison between
agreements — use
legal-reasoning. - Explaining what changed between marked-up versions — use
summarize-redlines. - A key terms table across documents — use
compare-key-terms; an obligations tracker — useextract-obligations-and-deadlines. - A written memorandum of advice — use
draft-legal-memo. - A defined terms or cross-reference audit of the finished draft — use
check-defined-termsorcheck-cross-references. - Pure file format work with no precedent involved — use
docxorpdf.
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.
- 2d ago First seen · 237 lines · 151 tokens per session scan A 0c4796c8128f
draft-new-contract is a skill published in the GitHub repository microsoft/cat-agent-skills (101 stars, last pushed yesterday), licensed MIT. It adds 151 tokens to every session and 3,172 once invoked, about $0.0006 per session on Opus 5.5. 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 2026-11-06.
draft-new-contract compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| draft-new-contractmicrosoft/cat-agent-skills | Skill | 101 | yesterday | MIT | A | 151 | Claude Code, Codex |
| html-ppt-zhangzara-stencil-tabletnexu-io/open-design | Skill | 107k | today | Apache-2.0 | A | 49 | Claude Code, Codex |
| treatment-plansK-Dense-AI/scientific-agent-skills | Skill | 56k | 6d ago | MIT | A | 62 | Claude Code, Codex |
| legal-holdanthropics/claude-for-legal | Skill | 9.9k | yesterday | Apache-2.0 | A | 68 | Claude Code |
| handbook-updatesanthropics/claude-for-legal | Skill | 9.9k | yesterday | Apache-2.0 | A | 52 | Claude Code |
| terms-of-servicegenspark-ai/genoffice | Skill | 12k | today | Apache-2.0 | A | 32 | Claude Code, Codex |
| software-copyright-materialsFokkyp/SoftwareCopyright-Skill | Skill | 6.3k | 16d ago | AGPL-3.0 | A | 98 | Claude Code, Codex |