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 agentmods add skills/wangyuecheng1106/magic-mouse/cursor-skillnpx skills add WangYuecheng1106/Magic-Mouse --skill cursor-skillgit clone --depth 1 https://github.com/WangYuecheng1106/Magic-MouseWrote 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/wangyuecheng1106/magic-mouse/cursor-skill)<a href="https://agentmods.dev/skills/wangyuecheng1106/magic-mouse/cursor-skill"><img src="https://agentmods.dev/badge/skills/wangyuecheng1106/magic-mouse/cursor-skill.svg" alt="Measured on agentmods" 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 | $0.00000 | $0.03511 |
| Opus 5 | $0.00000 | $0.01755 |
| Sonnet 5 | $0.00000 | $0.00702 |
| Haiku 4.5 | $0.00000 | $0.00351 |
Grade C, and why
cursor-skill scanned grade C 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 5d 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.
Recursive force deletehighDestructive command
rm -rf with a variable or a broad path is one typo away from removing the wrong tree.
- NEVER call `mm_shell_run` with destructive commands (`rm -rf`, `format`, How it starts
The opening of the file, as written. The whole thing — 356 lines — stays where its author put it; the contents beside it link to each section on GitHub.
name: magic-skill description: >- Operate the user's computer through Magic Mouse — a desktop pet that is the physical body for AI-driven screen interaction. Take screenshots, reason about what you see, click/type, and verify results — a full computer-use loop enhanced with Agent-Computer Interface (ACI), hierarchical planning, and dual-memory experience learning (Agent-S + Codex Computer Use hybrid). Use when the user asks to perform actions OUTSIDE Cursor (control other desktop apps, fill forms, grab screenshots, paste content, automate desktop workflows, close popups/ads), or when the user mentions Magic Mouse, desktop pet, virtual mouse, "桌宠", "让 Cursor 操作我的电脑", or "用魔法鼠标".
Magic Skill — Computer Use via Magic Mouse (CursorDeskPet)
You are CursorDeskPet, an AI desktop pet with full Computer Use ability: observe the screen, understand interfaces, manipulate mouse and keyboard, launch/switch/close apps — all through the Magic Mouse runtime on the user's desktop.
Your service target is Cursor (and indirectly the user). You translate natural-language tasks into safe, explainable, undoable GUI action sequences, helping Cursor achieve Codex-level desktop automation.
Magic Mouse must be running (check with mm_status first).
Design heritage (dual-project hybrid)
| Capability | Source | How it lands in Magic Skill |
|---|---|---|
| Pixel-level mouse & keyboard ops | open-codex-computer-use | Screenshot + element positioning + click / type / hotkey |
| App takeover (@App naming) | open-codex-computer-use | launch / switch_window primitives + app allowlist |
| Safety confirmation mechanism | open-codex-computer-use | ask_user + destructive-op intercept |
| Hierarchical task planning | Agent-S | Task → subtask → atomic action decomposition |
| Agent-Computer Interface (ACI) | Agent-S | Structured action primitives + UI tree element_id |
| Narrative + episodic memory | Agent-S | Task-template memory + step-level experience memory |
| Self-evaluation & continuous learning | Agent-S | Post-subtask reflection → memory update |
What ships with it
14 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.
- hooks/cursor-hooks.json 857 B
- hooks/magic-mouse-state.mjs 813 B runs code
- package.json 174 B
- reference/examples.md 3.4 KB
- reference/tools.md 4.9 KB
- server/package.json 812 B
- server/src/auth.ts 993 B runs code
- server/src/client.ts 5.8 KB runs code
- server/src/index.ts 2.7 KB runs code
- server/src/memory.ts 4.6 KB runs code
- server/src/tools.ts 18 KB runs code
- server/tsconfig.json 469 B
- server/tsup.config.ts 537 B runs code
- templates/mcp.json.snippet 146 B
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.
- 5d ago First seen · 356 lines · 0 tokens per session scan C 9a2319fad17d
cursor-skill is a skill published in the GitHub repository WangYuecheng1106/Magic-Mouse (1 stars, last pushed 1mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 3,511 tokens. A static security scan graded it C with 1 finding (recursive force delete). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
playwright-recording
Record browser interactions as video using Playwright. Use for capturing demo videos, app walkthroughs, and UI flows for Remotion videos. Triggers include recording a demo, capturing browser video, screen recording a website, or creating walkthrough footage.
website-to-video
Capture a general website/URL and turn it into a HyperFrames video (site tour, showcase, or social clip from the site's own visuals). Uses headless Chrome screenshots + brand assets. Use when intent is general — portfolio/blog/landing-page showcase or social clip from the site. NOT for: product/SaaS launch or promo (→…
qa/e2e-playwright
Playwright E2E 测试完整方法论，涵盖项目初始化、Page Object Model、认证复用、API Mock、视觉回归、多浏览器测试、CI 集成和调试技巧.
playwright-skill
Complete browser automation with Playwright. Auto-detects dev servers, writes clean test scripts to /tmp. Test pages, fill forms, take screenshots, check responsive design, validate UX, test login flows, check links, automate any browser task. Use when user wants to test websites, automate browser interactions…
chrome-devtools
Browser debugging, performance profiling, and automation via Chrome DevTools MCP. Use when user says "debug this page", "take a screenshot", "check network requests", "profile performance", "inspect console errors", or "analyze page load". Do NOT use for full E2E test suites (use playwright-skill) or non-browser…
agent-browser
Browser automation CLI for AI agents. Use when the user needs to interact with websites, including navigating pages, filling forms, clicking buttons, taking screenshots, extracting data, testing web apps, or automating any browser task. Triggers include requests to "open a website", "fill out a form", "click a…