Claude Desktop 安装指南
支持本地 MCP 配置的桌面客户端。
MCP Servers
AI-Powered Civil 3D Modeling
Automation-ready anti-detect browser workspace with isolated profiles, REST API, MCP, Docker, and Playwright control.
MCP server for Kdenlive — wraps kdenlive-api for Claude Code and other LLM agents
Python implementation for building Users Management Agent with MCP tools and MCP server
An MCP server for reading, rendering, and searching PDF files. Built with PyMuPDF and PyMuPDF4LLM.
An MCP server that connects Claude Code to Google's Gemini 3.1, unlocking capabilities that complement Claude's strengths.
Opencode-Raven keeps noisy tool work behind a focused Raven agent, so your main model gets compact answers instead of raw search results, docs pages, GitHub examples, MCP output, or large MCP schemas.
GUI + MCP control plane for your local dev servers. Humans click, agents automate the same daemon.
A lightweight Linear MCP for Claude Code. One tool, seven actions.
Query OpenAI models directly from Claude using MCP protocol.
Autonomous DFIR agent — SANS SIFT Workstation MCP server for incident response
轻便的人机共读MCP
Agent Skills
Linear Chain-of-Thought anchors on whatever it says first. Tree-of-Thought widens the search but still walks a single shared context, so the anchoring persists across branches.
Repeatable agents-plus-code workflows, packaged as one skill, stamped into any repo. Deterministic Python owns the graph; coding agents are bounded nodes inside it.
Use any agent to create stunning, interactive presentations
video-spec-builder —— 把我想做个视频逼成一份精确到秒的分镜脚本 video-spec.md,交给 HyperFrames 渲染。一条命令装到 Claude Code / Cursor / Codex:npx skills add feicaiclub/video-spec-builder
VulnHunter is an open-source, that applies proactive, attacker-first analysis directly to source code.
CAR is a meta-harness for coding agents (Codex, Hermes, OpenCode, OMP, anything ACP). It is not a coding agent itself. You bring the plan and the agents; CAR coordinates the long-running execution.
1. 对话 skill 模式：在聊天中按 SKILL.md 约束执行单轮改写；安装本仓库作为 skill 后即可直接使用。 2. 脚本 API 模式：通过 scripts/run_aigc_round.py 执行单轮分段处理，并按需调用外部 OpenAI 兼容接口。 3. Web 模式：启动本地后端和前端页面，在浏览器中完成上传、配置、执行和导出。 4. app：安装app可直接使用
得到 APP 课程下载工具，可在终端查看文章内容，可生成 PDF，音频文件，markdown 文稿，可下载电子书。可结合 openclaw +skill 等使用
面向 Web 逆向分析与接口签名还原场景的 Skill，围绕 camoufox-reverse MCP 构建单一工作流：先用 Camoufox 反检测浏览器完成网络捕获、源码定位、Hook 调试与反检测验证，再按需落地到 Node.js 或 Python 算法还原与自动化调用。
Slide decks, landing pages, interactive prototypes, interactive explainers, animations, design systems, and more — powered by structured design thinking and built-in quality guardrails.
将本仓库中的 skills/audit-skills/ 整个目录放置到 claude 或 codex 可识别的 .skills/ 目录下；能力范围覆盖 Java、.NET 与 PHP 审计：
🦀 Crabwalk 🦀 Real-time companion monitor for OpenClaw agents.