compuseTop-level cross-platform computer-use skill that bundles standalone macOS, Windows, and Linux runtimes with zero local Claude dependency and selects the corr...
Install via ClawdBot CLI:
clawdbot install wimi321/compuseGrade Fair — based on market validation, documentation quality, package completeness, maintenance status, and authenticity signals.
Potentially destructive shell commands in tool definitions
rm -rf /Calls external URL not in known-safe list
https://github.com/wimi321/macos-computer-use-skillAudited Apr 17, 2026 · audit v1.0
Generated Oct 7, 2026
A software QA team needs to run GUI regression tests on macOS, Windows, and Linux build agents from a single skill. Compuse installs all three platform payloads and selects the correct runtime project based on the host OS. Testers avoid maintaining separate toolchains per platform while still respecting native limitations.
An internal IT team uses Compuse to automate repetitive desktop tasks such as launching applications, entering credentials, and inspecting window focus across mixed-OS employee fleets. The skill's MCP layer enables action gating and logging for compliance. Support agents can trigger scripts through a shared multi-OS framework.
A fintech operations group prototypes robotic process automation for legacy desktop apps across macOS and Windows workstations. Compuse provides portable computer-use primitives without relying on a local Claude install, and its per-platform bootstrap installs dependencies on first run. Teams validate the macOS path first and plan Windows/Linux rollouts.
A mobile/desktop device lab administrator needs one skill to drive UI interactions on leased macOS, Windows, and Linux machines. Compuse resolves the active platform and builds only the required runtime, reducing image size per host. Linux X11 support is used for screenshot and synthetic input testing, with Wayland limitations documented.
An AI research team uses Compuse to programmatically record GUI interaction traces across operating systems for model training. The bundled runtimes let them collect typing, focus, and screenshot data uniformly. The skill's explicit platform validation notes help them label which host data is real-device verified versus pending validation.
Compuse is distributed as a free open skill with bundled macOS, Windows, and Linux runtimes. Enterprises pay for validated Windows/Linux builds, security reviews, and priority support for mission-critical automation. Revenue comes from annual support contracts and validated platform certifications.
A provider hosts Compuse-equipped virtual desktop workers and sells task automation as a monthly service. Customers pay per concurrent bot or per completed workflow across any supported OS. The provider handles patching, platform selection, and MCP-level action gating.
Consultancies package Compuse into client-specific computer-use automations, covering platform selection, bootstrap setup, and compliance guardrails. They charge project-based fees for initial deployment and retainers for ongoing maintenance across OS updates. Upsell includes training and custom MCP tool development.
💬 Integration Tip
Build only the platform project returned by `current-project.sh` to avoid pulling unnecessary runtime dependencies, and always confirm the host's real-device validation status (macOS verified, Windows/Linux pending) before claiming a workflow is production-tested.
Scored Oct 7, 2026
Guide for creating high-quality MCP (Model Context Protocol) servers that enable LLMs to interact with external services through well-designed tools. Use when building MCP servers to integrate external APIs or services, whether in Python (FastMCP) or Node/TypeScript (MCP SDK).
通过 macOS Accessibility API 控制小红书(rednote)App。支持私信、评论回复、视频评论读取、搜索、点赞、收藏、作者数据等。需要 Mac + rednote App 可见 + Terminal 辅助功能权限。
AI-first security intelligence with LLM-powered intent analysis. 743+ findings from 361+ skill audits, 25 pattern rules, 22 attack classes.
OpenAMC 本地 MCP 命令调用技能。通过标准化 CLI 接口,允许 AI Agent 主动向本地 MCP Server 发送查询指令以获取全球金融数据。覆盖 A股/港股(AKShare)、美股(yfinance)、宏观经济(FRED/IMF/EconDB)、美国国会(Congress)及衍生品、外汇、大宗...
Build backend AI with Vercel AI SDK v6 stable. Covers Output API (replaces generateObject/streamObject), speech synthesis, transcription, embeddings, MCP tools with security guidance. Includes v4→v5 migration and 15 error solutions with workarounds. Use when: implementing AI SDK v5/v6, migrating versions, troubleshooting AI_APICallError, Workers startup issues, Output API errors, Gemini caching issues, Anthropic tool errors, MCP tools, or stream resumption failures.
Interact with the BNB Chain Model Context Protocol (MCP) server. Use to query DeFi data, get token prices, search documentation, fetch git diffs, and retrieve smart contract source code on BNB Chain.